32,679 titles preserved for the world!
175 in Aug 2016 — 150 in Sep 2016 — More...
|DP · Blog · Forums||· Register · Help|
(Version 3.8; last updated May 20, 2010)
Note that there are additional references at the end of this document under See also ...
The purpose of post-processing is to massage the pages worked on in the rounds into a final etext for uploading to Project Gutenberg (PG). On its journey through multiple proofreading and formatting rounds, the text may have been worked on by hundreds of volunteers. The post-processor must standardize the formatting of the book and adjust it to comply with Project Gutenberg's requirements. They must also deal with any detectable mistakes that have survived all proofreading and formatting rounds. The ultimate goal of post-processing is to create a consistently formatted etext, which contains as few errors as possible and which accurately reflects the intentions of the author. A plain-text version is always needed (a .txt file), but many projects now also require other formats. Don't be put off by this—there are people who can help with them if you don't want to do them yourself.
Post-processors require more experience than ordinary proofreaders. Since they are preparing the text for uploading to Project Gutenberg, they make choices and decisions about the layout and look of the text. Because of this, post-processing is usually only available for volunteers who have completed a number of pages in F1. Refer to the chart on your Activity Hub to see if you already have access. If you do not, clicking on the icon will take you to a statement of whether you meet the requirements, and, if so, a button to request access (this is automatically granted).
If you are not yet eligible, but have a reason for wanting to post-process (special language skills are a common basis for exceptions), please request access by sending an email to firstname.lastname@example.org.
This FAQ contains a lot of information, particularly in the Help section. The Post-Processing Forum also has some helpful sticky threads, especially the No Dumb Questions topic which is the best place to post new questions. For faster help, try using Jabber to visit the DP chat room—there's usually someone around who can point you in the right direction (all time zones covered!).
You can also get PP mentoring help. PG produces its own guidelines. Finally, a strongly recommended step is to look at existing PG books, especially if you can find a book similar to the one you are about to post-process (same author, related topic, etc.).
To create and check an HTML version you will also need to use an online HTML validator and CSS validator, Tidy, and a link checker. If your project contains illustrations, you will need an image editor.
There are other useful programs available which are not essential, but which can be extremely useful and will usually save you a lot of time. Some have Gutcheck built in, such as Guiguts, for which there is a lot of support, a PP Checklist using Guiguts, and a PPing tutorial.
If using a Mac, see Post-Processing on the Mac.
For your first project, it's best to pick a fiction book with a relatively small number of pages (less than 200 or so). Here's why:
Many post-processors did not follow this rule to start with, and have turned out all right anyway. But it's a reasonable guide.
There are three good ways to find a book for post-processing:
Note: There is no commitment in volunteering for a book. If you wish to stop post-processing it for any reason, it's best to contact the person who assigned it to you, so they can pass it on appropriately. They will not twist your arm or criticize you for not finishing. Think of it as allowing someone else the chance to work on it, and freeing up your own valuable time for another project!
Download your chosen text by going to its project page, scrolling to the bottom of the page, and selecting "Download Zipped Text". Do not select "Download Zipped TEI Text" (a different encoding) unless you know what TEI is and want to work with it. The plain text is the version that you need.
Scroll through the whole text to see if there are any difficulties, like footnotes, poetry, foreign languages, dialects, and tables in it. This way, you will know what you will be dealing with before you commit to the project. If you see any of these items, you might want to pick a different project. But if you think you can handle it, give it a try! There's always lots of help available.
Check the project thread for your book's title to see what proofreaders have been saying about it. Again, this can alert you to issues that might make the work more difficult than you had realized.
If you decide you want to work on the book, and the book was not already checked out to you by the project manager, then go to the project page, scroll down and select the "Check Out Book" button. Make sure that the book appears as checked out to you, or you might end up working for hours on it only to find that someone else has checked it out and submitted it! Your post-processing choice will appear near the middle of the post-processing page.
Note: Occasionally, perfectly lovely books appear in the PP Pool (at the bottom of the post-processing page). If you wish to take one of these to work on, click on its title on that page, perform the above checks (downloading text, reviewing the project discussion, etc.) and if you still want to go ahead, scroll to the bottom of the project page to the "Check Out Book" button. Click it, double-check that the book has been assigned to you, and start work.
It's very difficult to answer this question in advance. The time that a book will take to complete depends on three factors:
It can vary from several hours to several days. Some especially difficult works can take weeks (or more!) to complete. Remember to save your work often, using a new filename each time, so that if you make a mistake, you can easily recover. Take it at your own pace—you will be the last person going through this book in detail before its posting (although two other people will verify your work).
Try not to feel discouraged if it seems like it takes a long time to complete an "easy" book. Concentrate on learning the process of post-processing, familiarizing yourself with any tools you might be using, and doing a quality job, rather than on working quickly. You will speed up naturally with practice.
If you realize that the project you've chosen is too complicated for you, or if you find yourself short of time, it is perfectly okay to return the project to the PP pool using the "Return to Available" button on the project page. You can look for an easier project straight away or take another one when you have more time.
To return a project, find the title of the book you are working on on your PP page and click on its title. That shows you the specific project page for that text and, scrolling to the bottom, you will find an option for Return to Available. Click that button to go to the upload page. To put the project back into the Post-Processing Pool, leave everything blank and click the Return project button.
If you have done considerable work on the project, you can upload a partially finished project to the PP pool for another post-processor to complete. After choosing the Return to Available option, when the upload page opens, you can upload a zipped folder and leave a comment for the next person.
Many post-processing tasks can be automated using tools designed to minimize the complexity and repetition of such jobs. Please refer to the software-specific tutorials or user guides to find out how to use these utilities effectively.
Read the project comments and project thread for your post-processing project. If the proofreaders found anything of concern, make a note of it for special attention while processing the text. You will also need to make sure you follow the project manager's instructions for the text. Many request an HTML version as well as the text version.
You may like to put notes about your progress in the "Post-Processor's Comments" box which can be found on the project page of your project, towards the bottom. These comments are visible to those who need to see them. They can act as a "To Do" list, or notes on points to watch out for from proofreader comments, and are particularly useful if you have to take a break from post-processing for a short while, so you can start working with your project right where you left off. Also, if you leave DP your notes could help another volunteer to take over where you have left off.
There is a Guiguts PP Process Checklist that may be useful if you use Guiguts. Some post-processors make their own checklist, some in a spreadsheet format, some at the top of the text they are working with (but if you use this option remember to remove your list before uploading the project for verification).
Check through the text page by page, opening the corresponding page scan in your image viewer. You'll quickly notice unmarked poems or block quotes this way. Check for missing pages (rare, but it does happen) and illustrations. If the project has problems like missing pages, it would be nice if you could go to the project page and state the issue in the comments in order to make others aware of it.
Run a search for * to find notes left by proofreaders/formatters to make you aware of their questions/solutions and potential problems.
Make sure that the /* */, /# #/, etc. tags are balanced. Be sure that any poetry is in the correct markup to save messes later. This is a good time to check each poem is indented correctly or has the relative indents correctly added. Every <i> tag needs a closing and properly placed </i> tag and so on. You may wish to change some formatting tags to markup specific for your post-processing tools (e.g. /p p/, /f f/); check your tool's manual for details. Some PMs may request particular markup in the rounds. Also, check any markup that ranges over a page break and make sure it will still result in the desired formatting (usually by deleting all but the first and last markers for a particular section).
All of these markups normally should have a blank line before the opening tag and a blank line after the closing tag. They should be on a new line with no other text, unless your post-processing tool allows it.
When formatting the title page, you have a bit of leeway. You can adjust the pieces a bit if you like: for example, you could move the author's name directly under the "by". Relative indenting is not required, but can be added if you wish.
Do block indent a consistent amount (from one to four spaces) if there are consecutive lines that should not be rejoined later in the process. The space is a flag in many text readers to preserve the given line endings should the general text need to be rewrapped to different margins.
For the table of contents and list of illustrations, please retain the page numbers. Line up the chapter titles and page numbers to make it look neat and easy to read. Copying the original format of the table of contents usually works fairly well. Leave all the original information on the title page, including the edition, year of publication and any copyright notice (unless this is a reprint—check with the project manager if in doubt). It is better to keep as much information as possible than to try to find it once the book has been posted for years.
You will need to rejoin footnotes split across pages. Then, in the plain-text version, you can put the footnote after the paragraph it refers to or at the end of the chapter or section. Make sure that the number/letter/symbol in the text matches the tag in the note itself. In-line footnotes (footnote within a line of text) are discouraged even when extremely short.
Consider using end-of-paragraph footnotes if the footnotes are short, unique, and not common. Use end-of-section or chapter footnotes for longer footnotes (such as those that have poetry or block quotes), or those that have multiple references in the text for one footnote. Whichever you choose, be consistent within the work. Use all end-of-paragraph footnotes or end-of-section/chapter footnotes within one work. Don't switch back and forth.
For the HTML version, they can be moved to the end of the chapter or section or to the end of the project. They also need to be hyperlinked. Most of the post-processing tools will do this automatically. Refer to the tutorials, guides, or manual for your software, to find out how.
The preferred method is to renumber the footnotes so that each one in the book has a unique number, alphabetic letter, or Roman numeral to make it easier for the reader to search the text. Alphabetic letters and Roman numerals are not recommended for more than 20–30 footnotes as they become hard to read/distinguish. There may be some projects where you may prefer to retain the numbering as in the original publication.
Move each illustration tag to an appropriate paragraph break. Some post-processors like to have them just before, or after, the text they illustrate. Others prefer to place them at the end of the chapter, not wishing to interrupt the flow of the text. Do whatever you think is right for your book.
Note: Keep illustration markers in the plain-text version, in case people want to refer to the HTML version later. Please do not delete them unless requested in the project comments and/or discussion.
If you do not want to produce an HTML version, but your book has pictures, post in the HTML pool, where you can enlist someone to generate the HTML and pass it back to you for uploading. HTML versions are required for every book produced at DP with pictures (even if the project manager does not request it).
Remove page separators, checking either side of them to see if the next page requires a blank line, is a section or chapter, or needs to be continuous text. You can rejoin words split across pages at this point if you haven't done so earlier.
Even if it looks like it's going to be a pain, spellchecking is always needed. Texts written before spelling was regularized might be the only reasonable exception, but even for those spellchecking is often useful. Even books with dialect or other deliberate non-standard spelling can be spellchecked. You may want to leave this step until later in your checklist, and/or repeat the spellcheck whenever you type in new information, including a transcriber's note.
These may be run by separate tools or by your main post-processing program. Refer to the manual or tutorial for the toolset you are using, or ask in the Post-Processing Forum.
Examples include "smart" programs which can check for he/be irregularities, or regexes (a form of search) which flag unusual letter combinations, such as "tb" (possible scanno for "th") or "rn" (for "m").
Various regex-searches are available and some tools will run these as a set, through your usual search-and-replace box—again, check the manual/tutorial for the software you're using. Otherwise, have a look at the Regular Expression Clinic for more information and help.
A great formatting check to run is the regex \n\n\n which catches all chapter and section spacing allowing you to confirm their consistency, as well as finding any extra line breaks between paragraphs—especially common after block quotes or poetry. It's a good idea to run this again on the text version, after you've removed markup such as /**/ and /##/. (See below.)
Time to rewrap. PG advises to keep the HTML version as close as possible to the text version, so some post-processors will use the rewrapped text version for the basis of their HTML version of the ebook. Some however prefer to use the text before rewrapping to avoid having to adjust the text version a second time after formatting tags have been converted—see Creating a plain text version.
Did you see any poetry, tables, etc.? If not, rewrapping the lines should be easy. You will need to rewrap the lines to around 65–75 characters in length. (See PG's recommendation for line length.) Each program has a different way of doing this, and you will have to find the way that works best for you. Read the manual or instruction book for your utility.
If you found poetry, tables, etc. care needs to be taken when rewrapping that line endings are preserved as intended, and that they are block indented from at least one to four character spaces to prevent rewrapping in future versions of your texts.
If worst comes to worst and you cannot find an easy way to rewrap the lines, find and replace all line breaks with spaces, count any line to find approximately where 60–72 characters falls, and insert line breaks manually at this point. It's painful, but it works. (Be grateful that you chose a book with a low page count!) Alternatively, type a line like this:
at the top of your text and use this as your guide. However, manually rewrapping in this way should not be necessary.
Once your text is suitably rewrapped, remove any end of line spaces. (Again, use the post-processing software wherever possible! All current tools include this task.)
The Gutcheck tool was written specifically to pick out many of the most common problems with PG texts. It is probably the single most important check you will perform. Follow the instructions with your post-processing software. If you are not using a post-processing-specific tool, you can download Gutcheck from here, and run it according to the instructions given there. Either run the check initially with all options turned on, or run each check individually, but make sure not to skip any. Check every potential problem that it brings to your attention. Not all Gutcheck "flags" are genuine errors (for example, it may report short lines where the text contains poetry or a table), but each must be looked into and corrected if necessary. Continue to run Gutcheck after each series of corrections until it doesn't flag any more "true" errors.
If you do not want to download Gutcheck, use Project Gutenberg's online gutcheck.
If you make any changes to the text it is a good idea to include a Transcriber's Note. Sometimes these are quite simple:
Transcriber's Note: Punctuation has been normalized.
Sometimes the writing of the Transcriber's Note is not always as straightforward as it might seem. Some suggest using wording such as "obvious errors have been corrected" but others say that what is obvious to one person may not be obvious to another. Also, correcting what someone might think is an obvious error, may in fact be correct spelling/phrasing for the time the book was written.
A useful general one, especially for older, less regular texts, is:
Transcriber's Note: All apparent printer's errors retained.
This one stops the PG whitewashers from getting long errata requests to "fix" your text. It is not, however, an excuse for leaving in bad OCR, scannos, or similar detectable problems that are wrong in comparison to the page scan.
Sometimes the notes can be quite lengthy.
Transcriber's Notes: Page 13, "10,00 troops" changed to "10,000 troops." (We fought 10,000 troops at St Germaine.) Page 27, "Faw-cett" changed to "Fawcett". (Major Fawcett dictated the memo.) etc. etc.
While we don't retain the individual page numbers in the text version, this gives the reader an idea of where it is in the book. The reader can search for the text you have included in the parentheses to find the exact location of your edit.
In the HTML version, the use of "hover" or "inserted" tags is a good way to shrink your list of changes while still maintaining the integrity of the original. Check the post-processing forum for ways of doing this, or follow the instructions here: CSS Cookbook—Corrections.
Many post-processors do fix what appear to be printer's "errers" (such as changing "errers" to "errors"). Do not modernize or switch the spelling from British English to American English or the other way around however. We are preserving history, not improving it.
Some put shorter notes, or ones that apply to the whole text in a general way, at the start of the book (before the title page), and longer lists at the end of the book (after any index or footnotes).
Transcriber's Notes are optional, but can help the reader's understanding of how you've processed the text. It's up to you how much or how little you note. If in doubt, talk to other post-processors in the forums, Jabber, or by PM about how they've handled various situations.
At the end of the above process, you have a processed book which contains HTML markup, as well as DP tags like [Footnote] or <tb>. Save a copy of this "dual" purpose file, calling it something like <name-backup.txt>.
PG will accept alternatives to the following. The important thing is to be consistent throughout your book.
If you want to make your book available for smoothreading, now's the time.
Mac and *nix users need to change line endings to CR/LF.
An extra pair of eyes is always helpful in finding things you might have overlooked in the text. Smoothreading is an option available to all post-processors and is generally done on a text version.
Save a new version of your book (such as <funnyname-smooth.txt>), then place this file into a zip folder.
Make sure that the file name contains some combination of
a-z, 0-9, -, _ and one . separating the filename from the extension
(no capital letters, no spaces, no special characters other than
those mentioned above). For example:
Go to the project page for your book. At the bottom of the page, you have three options: make the project available for smoothreading for one week, two weeks, or four weeks. Select the desired duration and upload the zip folder with the text file for smoothreading. You can provide comments about what to look for during proofreading, or to ask for attention in a particular section (this is very helpful in long texts). You might also like to advertise the availability of your book in the project thread, or in relevant team threads (see the Teams List for ideas).
Smoothreaders will mark possible errors in the text with [**description of query]. This is a standard format and should not be altered in your comments. When they finish, they will upload the smoothread project back to the project. At the end of the smoothreading period, you can download the smoothread versions from the bottom of the project page and search the text for [**. Not all [**comments] will be valid, just correct those that are. Make your corrections in the master file which still contains <i> markup (or else make each change in every version of the text that you have, e.g. plaintext and HTML).
While your book is being smoothread, why not start work on any other formats that are required, or begin fixing up any illustrations?
Go back to your marked-up copy. Use a copy of the marked-up file, named something like: <funnyname-htm.html>. Make sure you keep a version of the marked-up file for backup and reference.
Also see PG's guidelines.
These few simple changes will make our books look good on e-readers, without having to understand all the principles described in the Best Practices document.
Note that external links are generally not permitted, except for links to other ebooks within the PG site. If used, there must be a disclaimer at the beginning of the file to explain that links going outside of the document may not work for various reasons, for various people, at various times.
Create a new zip folder. Keep the filenames short, with only letters, numbers, hyphens, and/or underscores—no spaces or special characters like ?, #, $, etc. Filenames and directory names must be all lower case. Add into this archive your plaintext file, and any other formats that you've made. Any illustrations should already be stored inside an "images" folder, and this entire folder should be added to the zip archive. If you've been post-processing with Guiguts, add into the archive the .bin files for the plaintext version and HTML version if they are available—this is incredibly helpful for post-processing verifiers (PPVers) who also use Guiguts. They won't be uploaded to PG. Be sure not to include the page scans in your zip archive.
Your zip folder should look something like this:
Check there are no system files such as a thumbs.db in any of the folders.
Depending on your zip software, you may have to adjust its settings to "Save Relative Paths". This prevents the PPVer from getting extra (undesired) folders on their computers.
If you are using a Mac, you may need to "omit Finder files" too (leaves out invisible files).
Go to the project page for your book, and select Upload for Verification from the buttons at the bottom of the page. Include an email address in your post-processing comments section if you want email notification when the book is Posted! You can also use these comments to note any checks you've done or point out special features of the work which the PPVer should be alert to. Ensure that your site preferences with regard to post-processing credits are correct, as they are what will be used to credit you in the finished book. If you do not wish to be credited, or would like a different name to be used, please note this in the comments.
Your email address will not be displayed in the credits line, but can be used by the PPVer to give you feedback (if you request that option). If you do not request this, feedback will be sent via a private message on the site.
First, your book goes to an experienced post-processor for Post-Processing Verification (PPV). This person will carefully go over your work making sure that all of the requirements have been met, i.e., spellcheck has been done, images are correctly sized and formatted, it passes Gutcheck, the HTML is valid, etc. Sometimes a PPVer will request that a project be returned to you for further work. This does not mean you are a horrible post-processor. It probably just means that you missed a step or two of the process. An email or private message will accompany a return explaining why and what steps you can take to repair your file and usually offering assistance or suggesting where assistance can be obtained.
After your work has completed PPV, the PPVer uploads it to Project Gutenberg. There a friendly Gutenberg whitewasher (WWer) will make a final check of your work (and the PPVer's work) and add the Project Gutenberg boilerplate of names and legal information. Sometimes a WWer will have a question for you and that question may come through your PPVer.
Finally your project is posted on Project Gutenberg for the world to enjoy! Congratulations! After your project posts, you can expect to receive feedback if you have not already received feedback from your PPVer. This feedback will tell you the great things that you did along with any suggestions for improvements in future work. Feel free to contact your PPVer with any questions that you have about your project. If you do not receive feedback and your project has posted, please drop a line in the PPVer help requested thread and someone will look into it.
If you find an error after the book has posted, (really, it sometimes happens), send a note to your PPVer with the details. The PPVer will contact the WWer. If it has been posted more than a week or two, you can send an errata note to PG.
Once you've had several projects run through the PPV process, and have been granted the ability to upload your work directly to PG, you will be sent instructions by the PPV coordinator on how to proceed. Please also see the Guide to Direct Uploading (DU) and Posting to PG for more details.
Sometimes the content provider accidentally skips one or more pages when scanning. They usually check through afterwards for any missing pages, but don't rely on this—check for yourself, too. Occasionally the scan is present, but part or all of it is unreadable. First, attempt to contact the PM to get a better scan. If the PM is for some reason unable to get a good scan for you, there are other people who can try to get these pages for you. Find them on the Missing Pages Wiki. You can research here, looking at various library catalogues for Missing Page Finders and contacting them by PM if you find a copy of your book in their library. If no catalogues seem to have the book, log into the wiki and post the book's details and your username in the "Missing Pages" list.
If you do obtain additional pages, illustrations or just replacement images, please email email@example.com. Include the location of the images on dpscans (your page-finder can tell you this), the title and projectID of the project, and the project will be fixed for you. Please wait for confirmation before uploading your project to PPV or PG. This is very important for archiving purposes.
If you have multiple sections of a single book and you would like to have them "stitched together" for ease of post-processing, please email firstname.lastname@example.org. Working on one file is easier than doing your own "stitching" or working on separate files, and is especially helpful if you wish to retain the original page numbers in HTML, or if the png image names overlap.
NOTE: Multiple volumes of a book can be posted to PG separately if appropriate—they do not need this treatment. If in doubt, ask the PM or email email@example.com.
Every project must have a plaintext file (unless the project absolutely will not work in a plaintext form—e.g. a musical score.) However, there are other formats, which add information and value to the basic text. See PG's File Format FAQ.
This is the most common non-plaintext format requested or required for projects. If you are working on a text which is part of an überproject or is a periodical, you may find a style guide defined for you—check the ÜberProjects forum. Otherwise it is up to you to make the project consistent and readable. To produce HTML, you may wish to use tools that you are familiar with if you have done web-editing previously. The major post-processing tools such as Guiguts will also produce basic HTML which will just need some polishing to be valid and look attractive. There is also PG2HTML which will generate a very basic HTML version for you to work with.
Ask in the No Dumb Questions thread for post-processors, and/or HELP! HTML thread if you'd like more help with this. Many people have learned HTML for the first time here, as part of their post-processing, and it doesn't have to be terribly difficult!
Alternatively, you can post in the PPing HTML pool, giving a little information about your project. Other DP volunteers enjoy the process of making HTML and will be happy to produce an HTML file for you.
HTML is essential for projects with illustrations. It is also very useful for projects with many footnotes (because they can be hyperlinked back and forth, making the text more usable) or with different letters used from the Latin alphabet (such as Greek, which can be encoded so that any reader with an adequate font will see the Greek letters). Even if your project has none of these, many readers will find an HTML file more readable than the plaintext, and if possible, it's always worth producing one.
If you do an HTML version, make sure that the <title> tags contain
the phrase The Project Gutenberg eBook of <put the name of the book here>,
by <and the name of the author here>. So for example, if the project
you were doing was "A Christmas Carol" by Charles Dickens, you would make
sure the title tags looked like:
LaTeX is a markup language, entirely distinct from other types of DP formatting, but normally confined to projects containing a lot of math. See the section on LaTeX below, or post in the LaTeX Forum for Post-Processors for more help.
TEI is a form of markup which is used to generate plaintext, HTML and other formats by automatic converters. PG does not currently use the full TEI standard, but is evolving a subset. See PGTEI. As long as it is a valid PGTEI document, David Widger of the Whitewashing Team will take the .tei master file and autogenerate the plain text, HTML and PDF files, but, if you have generated these files, include them with the upload as they make running Gutcheck easier for the WhiteWasher.
Check the post-processing forum for more information about PGTEI such as the PGTEI primer and PGTEI. Alternatively, you can post-process using TEI but submit only the product of your own transformations (plaintext, HTML and other formats if desired). Such projects are likely to be posted more quickly, and will look more exactly as you specify. PG prefers all versions of the files.
PDF versions are usually accepted by PG only if an original master file is supplied so that making any changes if necessary is not troublesome—for example, projects formatted using PGTEI. It is very useful for certain projects (e.g. those involving LaTeX) which benefit from a fixed paginated layout and embedded specialized fonts, but is less helpful for other projects because it is difficult to make changes to the PDF once it's complete. Some PG readers do like this format.
Project Gutenberg will accept proprietary text formats such as .doc, but prefers not to. Issues of software compatibility and conversion arise more frequently with these formats than with simple plaintext or HTML. If you are considering an unusual format, you may wish to discuss it with a PP Mentor or the project manager.
A UTF-8 file can be produced if you require characters which are not in Latin-1 (the "usual" character-set used by DP in the proofreading interface and dropdowns.) This probably isn't useful if you have a single word with an œ, but if your project has a fair amount of Greek or other characters, a UTF-8 version will preserve the text most faithfully. Most post-processing tools have support for this—check the user guide or manual for more information. Ask in the post-processing forum for more help with this. Also ask if you think you will need to use UTF-16 ... it's not common and there may be a good alternative.
If you produce a UTF-8 text file, name it something like: projectname-utf8.txt when you upload for PPV. This will help the PPVer and is how the Whitewashers have requested such files be labelled. The WWers have a script to automagically convert UTF-8 characters to a reasonable ASCII equivalent, so that both formats will be posted in the final PG archive. If you think that this conversion process will not produce a readable or useful ASCII file, you can produce one of your own with all UTF-8 characters translated. Name that file projectname-ascii.txt. Make sure it really is ASCII and that no UTF-8 characters have slipped in! Remember to include both files when you upload for PPV.
If you are unsure how to represent a symbol, post, with a link to the page image containing the symbol, in the Post-Processing Forum. Asking there will net you a varied range of ideas about whether the problematic ink blob is in Latin-1, Unicode, or can be improvised using ASCII-art or represented in another fashion.
Some things to look out for when handling footnotes:
Many post-processors panic when they see sidenotes. This is usually the wrong reaction (though is sometimes justified).
The simplest case is when there are few sidenotes, usually only one per paragraph, and usually at the start of the paragraph. In this case you can just put them before the paragraph they refer to. In the plain text it is probably best to leave them inside the [Sidenote: Text of sidenote.] markup, so the reader can tell what they are. Some people like to have them as headings, leaving a blank line between the sidenote and the paragraph.
In HTML, it is probably best to float them off to the side. You can choose whether to put them in the margin, so that they don't interrupt the flow of the text, or whether you want them to stick into the text which will then flow around them. It's probably best to follow the original as much as you can. If you want them in the margin, you'll probably want to use a larger margin than usual in order to make room for them. It is probably easier to read the HTML version if you put all the sidenotes on the same side.
If there are lots of sidenotes, with many sidenotes per paragraph, the situation gets more complicated. Putting them all at the start of the paragraph will lose information. In most cases there is a definite place in the text that the sidenotes are connected to, and that's roughly where the sidenote should go.
In the plain text there are at least a couple of options. The first is to put each sidenote (still in its [Sidenote: Text of sidenote.] markup) at the start of the sentence to which it refers. This has the advantage that the text stays easy to read, but, if the sentences are long, the sidenotes may end up quite far from their referents. The second is to try to place the sidenotes more exactly, by putting them in the middle of sentences. This makes for a text that is much harder to read.
Sidenote placement is easier in the HTML version in that it is easy to get them closer to their referents. However, you should check them in several different browsers and at different browser window sizes, as it is very easy to get them overlaying each other so that they are illegible.
Many post-processors have a hard time with images at first. Images are the most common reason for PMs to require an HTML version, but if you don't feel comfortable with HTML or dealing with images, it's okay! Make use of the HTML pool—there are many people who don't like to do the text portion of a project.
Not frightened off? Good!
Some projects have only one or two images, like a frontispiece or an author portrait. The image files should be resized, so you can then display them at their actual resolution within the HTML. For example, if you want an image to appear 400px wide in the HTML you should save it at that size, rather than saving a larger image and using HTML code to display it smaller.
Other projects are heavily illustrated, and often the point of posting them is for making the illustrations available to a wide audience. Image-heavy projects should use small copies at lower resolution and color-depth (commonly called thumbnails) that link to larger, better quality illustrations. This allows people on dial-up connection to the internet to get a feel for the project without having to wait hours for the HTML to load. Check to see if your PM included high-resolution illustrations along with the project pngs. These are usually located at the very end of the images and often have the extension .jpg.
A good rule of thumb is a maximum of 400px in width or 600px in height for thumbnails and full page illustrations, and a maximum of 1200px in the larger direction for linked-to images. The illustrations should scale accordingly. If your largest image is a full page illustration that you have made 400px, then your emblems that show up above the chapter header should only be a fraction of that. Imagine that you are holding the book, about what fraction of the page is the illustration taking up and adjust your px number accordingly. If you need to decrease the file size further, or touch up the pictures in some way, please see the Guide to Image Processing.
If even the high-resolution images are too dark or corrupted, contact the PM for replacements. Sometimes the PM has given you the best the book had to offer. Your next option is to try the Missing Pages Wiki to see if someone can provide better scans of the images.
All images for use in your final HTML should be stored inside a folder called images within the project directory. Do a final check, when you've completed work on the HTML, to make sure that all images are used correctly within your page, and that you haven't included any temporary or redundant files.
For any questions or advice about any illustration related matter—contact the Illustrators Team.
As long as you make sure your rewrap markers are set correctly, post-processing poetry shouldn't be any different from producing a prose book. Make sure you save backup copies of your file regularly as you work—it will be much easier to recover from a formatting decision gone terribly wrong. Have a look at recently posted poetry books at PG for layout ideas. Some post-processing software has extra features for handling poetry—refer to the user guide or manual for more information.
Sometimes tables in your text will already have received the careful attention of a member of the Turn the Tables team, and be sized to fit within PG guidelines (ideally, less than 75 characters wide, or 80 if desperate). If you need help with a table, or have questions about the HTML formatting, post in the team topic or in the post-processing forum.
Greek will usually have been transliterated (converted to Latin letters) during proofreading. There are various ways to handle this.
In the plaintext you can leave the transliteration, commonly removing the [Greek: Transliterated text.] markup although you may wish to use another markup of your own, such as a +, and mention its use to indicate Greek in a Transcriber's Note. Or, if you have a significant amount of Greek or other unusual letters, you can produce a UTF-8 version, which will contain the original letters. Post in the forums if you'd like help with this. Faster help can often be obtained for short phrases via the DP Jabber chat room.
In the HTML, you can always use Greek letters (whether your file is saved as UTF-8 or not); if it's not UTF-8, then you'll need to encode the Greek letters using HTML entities. Either way it will display for the reader if they have a relevant font installed. It's a nice idea to enclose the Greek in a <span> which uses the transliteration as a "title" attribute—that way non-Greek readers can still access the word. Again, ask for help if you need it. There really are people who enjoy doing this!
Check the Language Skills List to ask for help or advice. If a native speaker hasn't been at DP for some weeks, or can't help with your particular problem, have a look at the Teams Forum or on the wiki teams list to see if there's a team for the language or relevant country or countries. Don't worry if there are few members or the forum hasn't been posted to in a while—your question might be all it takes to create a lively and helpful community discussion.
Keep the page numbers in the index, even in the plain text version. In the HTML version it's very easy to link numbers to page anchors—see the user guide or manual for your post-processing software. If you need a way to do this linking semi-automatically (you'll need to check that non-index numbers aren't being included!), then just ask in the Regular Expression Clinic. For further help, or formatting queries, try the Junkies, Index team.
These should be included as printed. There are two ways to handle these: you can leave the amendments up to the reader, or you can make corrections in the text, adding a Transcriber's Note that you've done so. Whichever you choose depends on the project and on you as post-processor. Just don't make silent corrections and don't leave the pages out.
A possible middle ground would be to include the page's content as printed in the plaintext version, and use a correction <span> in the HTML version to indicate that a change was made to the text—making the erratum amendment, but including the original text to pop-up when the change is highlighted/hovered over by the reader's mouse pointer. See the PP forum for more on this.
Don't Panic! Everyone who post-processes has done this. If they haven't, they will eventually.
If the book is quite recently posted to PG (in the last week or two), contact your PPVer and let them know the problem. They'll pass it on to the WhiteWasher who archived your book and can most easily fix it.
If the book posted a longer time ago, contact PG's errata. Say that you are the post-processor of the book, and include the PG text number, title, and author with a clear description of the problem and how to fix it. If you've checked the problem against the page images, mention that too.
See the ÜberProjects Forum for a list of the large multivolume projects that are likely to be seen on DP for quite some time.
See also the Proofreading Periodicals team.
Many periodicals have a standard style for the text and HTML versions that ensures a consistent look for the whole project. If the periodical is part of an überproject, check the ÜberProjects Forum for details.
Many people are put off proofreading, formatting, or post-processing periodicals because they are perceived as "hard" in some way. Canny post-processors will therefore quickly realize that mastering a periodical style will give them access to many entertaining projects with little competition. A Style Guide puts an end to those hours spent mulling over whether a heading should be marked with <h2> or <h3>. Periodicals often have longer pages than usual and may have adverts or other unusual formatting issues. These will all have been encountered previously, and recommended handling should be explained in the Überproject thread. If not, or if the explanation is unclear, post there for help.
An excellent source of inspiration will be the most recently posted issues of that periodical at PG—refer to these to help. Make sure you select ones which have also been posted by DP to ensure absolute consistency of style.
Many people are put off proofreading, formatting, or post-processing drama because it is perceived as "hard" in some way. Sometimes it actually is quite hard, for example, when written in sixteenth-century English with little attention paid to spelling or grammatical niceties. Mostly though drama is quite straightforward.
For all plays, check the Formatting Guidelines, and ensure your plain text version is in line with these.
Format character names as similarly as possible to the original text. If the text is metrical (written like poetry where line breaks are significant), check the /* */ markers carefully before doing any rewrapping (or consider checking through for rewrap sections, such as stage directions, by hand.)
If the project contains unclosed brackets, be aware that Gutcheck will have many false positives.
There are various ways to format plays in HTML; searching PG for recent postings may give some ideas, as will posting in the Post-Processing Forum. The Plays The Thing team can also offer help and advice. Ideally, make it look as much like the original text as is sensible.
Books with sections of music, or a short tune for a song sung in the text, or entirely about music, are regularly put into PG by DP. A simple way to post-process a book containing music is to include all scores as illustrations in the HTML. However, much more value can be added to the book by transcribing the music into a common notation format. This has three great advantages:
The wiki music guidelines contain a detailed discussion of the different music transcription programs, such as Lilypond, Finale, Sibelius, and so forth, all of which can produce sound and image files, as well as editable source files. The guidelines also contain information for PPers about different ways to present music in the HTML, along with a list of sample e-books containing music.
To obtain help with music transcription, simply post a message to the Music team thread, or send a private message to one of the volunteer music transcribers listed at the end of the Music Guidelines page.
LaTeX is an integrated collection of typesetting software. Though oriented toward mathematical and scientific content not easily representable in plain text or HTML, LaTeX can create beautifully formatted printed works of all types, featuring complex page layout, camera-quality printed output, and auto-generated indexes, tables of contents, cross-references (including hyper-linked PDF), and bibliographies.
LaTeX projects submitted to PG must include the LaTeX source as a single file, together with any illustrations (in an "images" subdirectory, as for non-LaTeX projects). Projects must be compilable with TeX Live, which the PG whitewasher will use to generate the uploaded PDF. The MiKTeX distribution for Windows may be assumed equivalent to TeX Live for formatting and post-processing.
The primary repositories of LaTeX information and advice at DP are:
You probably want to speak the language, or have a native speaker spellcheck/smoothread it. However, for some languages, there are few or no native speakers available on the site. These projects can be taken by people who are willing to put in the extra effort involved in dealing with a language that they do not speak. Check the Language Skills List to find who to ask for help or advice.
When you submit a Latin-1 version of your text there is no need to also produce an ASCII version as PG has the tools to easily make an ASCII version based on the Latin-1 text. However, if the ideal ASCII-version would be different from the result you get by making standard replacements like ü -> ue, é -> e, etc., you should produce an ASCII version yourself. Explain your reason for doing so when uploading for PPV, so they can pass that message on to PG.
As a post-processor you have a bit of freedom in choosing the best format for your text. For LOTE texts this may sometimes lead to decisions which would be unusual or even plainly incorrect in English. If you make such a decision, you might get lots of Gutcheck errors. If you have a good reason for your decision and if you have applied it consistently, you can ignore those errors. You might want to mention this decision in your upload notes. Example: For many languages it looks more natural to have spaces around em-dashes. It's perfectly fine to leave or insert them.
You should replace English markup words which appear in the final text with translations of those words in the main language of the text. e.g. Footnote/Fußnote/Apostille/Ootnotefay/Υποσημειωση/Voetnoot/Nota de rodapé.
Distributed Proofreaders was founded in 2000 by Charles Franks to support the digitization of Public Domain books. Originally conceived to assist Project Gutenberg (PG), Distributed Proofreaders (DP) is now the main source of PG e-books. In 2002, Distributed Proofreaders became an official PG site. In May 2006, Distributed Proofreaders became a separate legal entity and continues to maintain a strong relationship with PG.