The difference between \\c and \\hd is that, while \\c centers a line anywhere on the page, \\hd "looks ahead" to see that a heading fits on the page according to the rules of braille textbook format. When a runninghead is being used, the rules of textbook format require a blank line between the runninghead and a subsequent heading. Using $s \\hd [heading] accomplishes this. \\c doesn't automatically skip line two, nor does it check to see that the heading is not the last item on the page. When a major heading coincides with the start of a new print page, \\pph must be used instead of \\pp. See \\pph for details.

\\mh

Use \\mh for minor headings. Textbook format rules call for a blank line before, but not after, a minor heading. \\mh blocks text to cell 5 from the point it is entered until it is turned off by \\rt. Enter it as follows:

$s \\mh This is a minor heading, and, unlike a major heading, it can be so long that it doesn't all fit on one line, and you don't have to manually divide it, because no centering is involved \\rt$people

Like \\c, \\mh works anywhere on the page; it doesn't "look ahead" for space as \\hd does.

A note about headings

Basically, braille uses two levels of headings: major and minor. The Code has very specific guidelines about their placement on the page. These include when and where surrounding blank lines are required and when they are prohibited. Throughout this manual, when we show how to use headings, we show surrounding skip-line indicators according to how they are used in textbook format.

There are other formats, such as magazines, that use headings slightly differently. The \\hd and \\mh commands work just fine when paragraph indicators or <CR>s are used instead of skip-line indicators; it is up to you to decide which is appropriate to use.

The key to making levels of headings consistent is planning. Before beginning a transcribing project, look through the book and determine how many levels of headings there are. When there are more than two, decide whether combining two or more levels would hinder the braille reader's understanding. When it would not, use only two levels of headings. When it is absolutely necessary to show more than two levels, use special heading indicators, discussed in Section 8, Part 5.

\\doublespace and \\singlespace

The \\doublespace command puts a blank line after every line of braille. It takes effect at the point it is entered, and continues to operate until cancelled by \\singlespace. An overriding goal in braille transcribing is to save space; \\doublespace therefore has very limited applications. One recommended use is in transcribing material for the first school grade. Because single spacing is the default value for all other \\ format commands, there is no need to use \\singlespace anywhere else.

\\ib and \\if

Learner Level BEX Sections 6 and 9 discuss underlining in print and braille. There is no way to physically underline in braille; the braille italics sign is used instead. For this reason, we used i mnemonic commands in TranscriBEX, rather than u mnemonics. When you are transcribing something that is either underlined or italicized in print, use \\ib at the beginning of that text. Use \\if after the last underlined or italicized word. TranscriBEX then appropriately places the braille italics sign, dots 4-6. When three or fewer words are to be italicized, a single dot 4-6 is placed before each word. When four or more words are involved, two dots 4-6 are placed before the first word, and a single dot 4-6 is placed before the last word to be italicized.

Section 10a of English Braille and Rule IV of the Code discuss the use and placement of braille italics in detail. Also discussed are ways of combining italics and boldface, several different typefaces, underlining colored type, etc. Section 8 of this manual covers more advanced translator controls and special symbols which handle these more complex instances. The \\ib and \\if commands are the simplest to use and are appropriate in a majority of situations.

\\items and \\enditems

The \\items command signals a list of things (words, phrases, sentences), as opposed to text in standard paragraph form. Use a paragraph indicator to begin each item. Each paragraph indents to cell 1 with runover to cell 3. Use \\items as follows:

• $p Fred--Here's a list of things I'd like you to pick up on your way home from the set: \\items $p a pair of tap shoes--size 5 $p an umbrella $p a new gramophone $p a cane \\enditems $p I'll see you at six! Ginger

The paragraph indicator works differently after \\items has been used. \\enditems restores the operation of the paragraph indicator to its previous indent and runover values. The commands \\items and \\enditems must be used as a pair, or the effect of \\items continues indefinitely. Always place the \\item and \\enditem commands before the first paragraph or skip-line indicator you want them to affect.

\\tn and \\endtn

Literary format allows the transcriber to add explanatory notes to text without using a transcriber's note. Textbook format, however, requires a transcriber's note for any and all text added in the transcription process which does not appear in print. The beginning and end of a transcriber's note are signalled by a dot 6 followed by a dot 3. Use a <CR>, then \\tn at the beginning of a transcriber's note. This generates the first dot 6, dot 3 and sets indent to cell 7 and runover to cell 5. Use \\endtn at the end of a transcriber's note. This generates the second dot 6, dot 3 and resets indent to cell 3, runover to cell 1. Follow it with a paragraph indicator, skip-line indicator or <CR>.

Please note: This is one \\end command which always goes back to the indent and runover values for standard prose paragraphs. When you use a transcriber's note in a format with a different indent and runover, you must follow \\endtn with that format command. For example, to put a transcriber's note in a poem do this:

• \\poem [some poetry] <CR>
• \\tn [transcriber's note] \\endtn <CR>
• \\poem [some more poetry]

For information on the differences between "transcriber's notes" and "transcribing notes which appear in print," see Section 13: Notes on Notes.

\\newpage

The \\newpage command starts a new braille page. Use it first whenever you want to force the formatter to move to the beginning of a new page. This command is "built into" all long-lasting format commands, such as \\bookprelim. When you need to use it for particular applications which don't begin with a \\ format command, such as a dedications page, we tell you. Don't worry about using too many though; \\newpage is a "smart" command. It won't work when it is already at the start of a new braille page, so you won't get extra blank pages in your output.

\\nobreak#

Use the command \\nobreak# before a block of text that you don't want to divide between pages, such as one section of a table. (Section 15 discusses tables extensively.) \\nobreak# is one of those commands that requires adding an integer of your choosing. Count the number of braille lines which must remain together as a block of text. Enter the command \\nobreak#, where the number is the number of lines that must stay together. When there aren't that many lines left on the page, the braille-formatter starts a new braille page.

Be cautious in your use of this command. Don't make the number too big. Of course, any number bigger than 25 won't work at all, because 26 lines can't even fit on a whole page, much less part of one! If you used \\nobreak frequently, and/or used two-digit numbers, you would end up with a lot of blank space--which is not a good transcribing practice.

\\rt

The \\rt command is used to resume "regular text." It re-establishes indenting of paragraphs to cell three, runover to cell one. Use it whenever you want to return to standard prose paragraph indent and runover values. It must be used to signal the end of a minor heading (see \\mhwith, for example, or to return to prose after transcribing a poem with \\poem.

\\rt and \\endtn always return to the standard prose paragraph values, indent to cell 3 and runover to cell 1. This is the big difference between them and the other \\end commands. All other \\end commands return to the values for indent and runover established prior to the use of the \\ command and \\end command pair.

Section 7: Making Braille: The TranscriBEX Process

Part 1: Overview of the Steps

Section 6 discussed how to place TranscriBEX's \\ commands during print data entry. But how do you get from print to braille? That's what this Section is about. It takes you through each step in the TranscriBEX process, using Chapter 8 from The Wisconsin Garden Guide as an example.

This first Part is an outline of the steps in the process; we explain each step in more detail in Parts 3 through 7. You use the same steps, in the same order, every time you make the transition from print to grade 2 braille. Consistency is very important! This process may seem complex the first couple of times you go through it, but if you're consistent about following these steps, in order, it soon becomes almost second nature.

The first step is the print data entry; in our Wisconsin Garden Guide example, we've done this for you. It's the chapter called WGG SECT8 on your TranscriBEX disk. As we go through the rest of the steps, we'd like you to actually perform them on chapter WGG SECT8. You can then compare your results to ours along the way.

The second step is to Replace characters, using MAKE$ as your transformation chapter. This turns all the TranscriBEX \\ commands entered in your text into the more complex BEX $$ commands. You name your target chapter by adding the single dollar sign character to the source chapter name. We call these formatted, untranslated chapters "the $$ chapters."

The third step is translating the print text into grade 2 braille, using the Grade 2 translator. The $$ chapter created in the previous step is the source chapter. You name your target chapter by adding the single digit "2" to the source chapter name. So, for our sample chapter called WGG SECT8, the formatted $$ chapter is WGG SECT8$, and the formatted and translated chapter is WGG SECT8$#2.

Now your text is formatted, translated, wonderful, grade 2 braille! Or, at least you hope it is; to make sure before you actually emboss it, the next to last step is to print it to the braille previewer. This allows you to catch and correct mistakes without wasting time and braille paper. When you find mistakes, go back and Edit your original \\ chapter, and send it through MAKE$ and the Grade 2 translator again until you're satisfied.

When you're happy with how your text looks on the previewer, it's ready to be embossed. Whether you perform this last step on site, or send the disk elsewhere, the final result is formatted exactly the same way it is on the previewer.

Part 2: Setting Up a Braille Previewer

It's quite important to your understanding of the TranscriBEX process that you follow the example in this Section through all the steps. One of the tools you need to do this is a braille previewer. (You must have an 80-column card to use this feature of TranscriBEX.) Don't worry if you don't understand exactly what the previewer is for right now; we explain that in detail in Part 6. But to avoid having to reconfigure in the middle of following this example, take the time NOW to set up a configuration which includes a braille previewer.

Using our sample configuration

In Section 3, Part 3 of this manual, we gave a sample configuration dialogue, defining a Thiel brailler as Printer 2 and a Video Mode brailler as Printer 3. Both braillers have a carriage width of 41 and a form length of 25. When you use this configuration and print a grade 2 chapter to Printer 3, it appears on the monitor screen in screen braille exactly as it would appear on the Thiel in dots.

To follow our Wisconsin Garden Guide example, you can use our sample configuration from Section 3, or make up your own. Here are step-by-step directions for setting up your own braille previewer: When the configuration prompt says, "Enter printer slot," answer "3." (This is where your 80-column card is.) When the prompt says, "Enter printer class," answer B (for brailler). The prompt then says, "Enter Brailler code." Your answer here depends on whether you can see the 80- column screen display or not. If you can, answer "1". This is VIDEO MODE, a braille previewer that displays screen braille precisely as it will appear on the braille page, with a "cheat sheet" to help you check for errors. If you can't see the 80-column screen, you need Echo line review; answer "2" to the brailler code prompt. This is VOICE AND VIDEO MODE. This braille previewer allows you to output a screen's worth of braille; you can then use Echo line review to read it one line at a time. BEX User Level Section 6, Part 4 describes the Voice and Video Mode, and how to use it with Echo line review.

The next configuration prompts ask for carriage width and form length. Answer these prompts exactly the same way for the previewer as for the actual braille device you are using. When you Print a chapter to a previewer, TranscriBEX "thinks" it's really printing; it formats the chapter exactly the same way it does when you Print to an external device. It is up to you to tell TranscriBEX what its boundaries are. If you defined carriage width and form length differently for the previewer than for the external device, TranscriBEX wouldn't be able to accurately show you on the previewer what the final result would be on paper. When you give the previewer the same parameters as the actual device, TranscriBEX sends identically formatted text to each.

Again, don't worry if you don't understand exactly what you're doing when you set up a braille previewer. That will become clear as you follow our step-by-step example--which you're now ready to do!

Part 3: Entering and Proofreading Your Text

The first step in the TranscriBEX process is to enter your text. Type the text in exactly as it appears in print; that's the content. Enter TranscriBEX's \\ commands to give the text its "shape", or format. As promised, we've done this part for you in our example. Examine chapter WGG SECT8 in the Editor. Compare it to the physical print copy on pages 7-16 through 7-22.

In Sections 5 and 6, we talked about how some of the more basic format commands, like \\textbookformat, \\runninghead, and the paragraph indicator ( $p ), are used. Look for them in chapter WGG SECT8. Some of the other \\ commands in WGG SECT8 won't look familiar to you yet. We discuss them in later sections of this manual. For now, you'll just have to take our word for it that they work!

Proofreading

After you've done your data entry, carefully proofread your chapter. Match it against your print copy; make sure you haven't omitted a page, repeated a paragraph, or entered an incorrect page number. Check for typos. TranscriBEX recognizes anything beginning with \\ as a command. When you misspell a \\ command, TranscriBEX doesn't know what you want it to do; it gives you an error message. When you misspell words in your text, TranscriBEX can't translate them correctly.

Let's take a close look at chapter WGG SECT8. To follow this process, first Initialize (option I on the Starting menu) both sides of two flippy disks. (If you don't have flippy disks, three floppies work just as well; we talk about flippies throughout this process.) Copy chapter WGG SECT8 onto side 1 of one flippy disk. Label this side "Print \\ chapters" (or something similar). Examine your copy of WGG SECT8 in the Editor.

Some familiar commands

There's a whole series of commands at the very beginning of the chapter, all of which were introduced in Sections 5 and 6. \\textbookformat establishes that print page indicators are used throughout the text. \\setnumber!!" sets the value of the sequential braille page number. (Since we're only dealing with one chapter in this book, we picked this number out of a hat. It could be higher or lower, but it needs to be there.) \\pph 220 tells TranscriBEX to put print page indicator 220 in the text. You use \\pph instead of plain \\pp so that it doesn't interfere with the heading. A runninghead reading "Wisconsin Garden Guide" is established. The command \\newpage is used to make sure the chapter starts on a new braille page. Notice that there are spaces between all the \\ commands except \\runninghead [text] and \\newpage. The <CR> takes the place of the space there. It is important to follow this format exactly; if you used both a <CR> and a space between some \\ commands, they wouldn't work.

The next command is \\hd, for the major heading at the beginning of the chapter. Notice that the large number eight which appears at the beginning of the first paragraph has been moved to the beginning of the heading. This is allowed by the Code of Braille Textbook Formats and Techniques.

Some commands you don't know yet

A short poem by Robert Louis Stevenson follows the chapter title. The commands used to format the poem, \\poem, \\attrib, and \\endattrib, are discussed in Section 16. The command \\rt marks the end of the poem format and signals TranscriBEX to return to the standard paragraph indent and runover values established by \\textbookformat. The body of the text follows, with paragraph indicators marking the beginning of every paragraph. The print uses two columns on every page. This format is not used in braille; simply enter the text in the sequence it is written in.

Minor headings, paragraph headings, and braille italics

The next thing to notice is the treatment of minor headings. The command \\mh is used for the "Living Lawns" heading. Notice that a skip-line indicator is used before it, and a paragraph indicator after it. The placement of the heading is sufficient to alert the braille reader that this is a minor heading. Do not also attempt to show the change in typeface. In print, the heading "Grasses for Wisconsin lawns" appears in smaller, boldface type. It is not set off from the text. Braille italics, marked by the \\ib and \\if commands, are used for this lower level heading. The braille italics commands are also used around the words "individual plant" (in the second paragraph under "Living Lawns") because they are italicized in print to show emphasis.

Our mistakes

We purposely made a couple of mistakes in our sample chapter to give you practice in using the locating features present in the Editor. Please correct these mistakes in your copy of chapter WGG SECT8 before continuing. The first mistake is that \\pp 222 is entered twice. One should be \\pp 221. To find this mistake, position your cursor at the beginning of BEX page 1 and type control-L 222 control-A. When the cursor doesn't move and you hear a beep, it means 222 can't be located on that BEX page. Move to the next page (control-P 2 <CR>were and type control-L control-A. You don't need to enter the 222 again; BEX remembers what you're trying to locate.

When you've located the first 222 in your text, correct it to 221. Notice that the \\pp command appears EXACTLY where the print page break occurs--right in the middle of a sentence. This is required by the Code.

The second "mistake" we made in our data entry occurs on print page 223. Locate the beginning of print page 223, as described above. The second complete paragraph on page 223 was entered twice. Move your cursor to the beginning of the paragraph by typing control-A 2 control-P. Now delete the paragraph by typing control-D control-P. The first occurrence of the paragraph disappears; the second occurrence remains intact.

Reproducing mistakes in the text

The Code requires that text be entered EXACTLY as it appears in print. On page 221, under the paragraph heading "Establishing a new lawn," The Wisconsin Garden Guide refers once to "James C. Shroeder" and once to "Jim Schroeder." You have no way of knowing which is the correct spelling; your job is to faithfully reproduce the print.

Dashes and hyphens

There are a few other things you should notice on print page 221. The first paragraph which begins on that page, starting with "The fescues--" contains several dashes. Enter dashes in your text as two hyphens. The translator puts a special control character after dashes and hyphens so that lines can break there.

Extracts from other sources

At the end of print page 221, a lengthy extract from another source begins. In print, the entire extract is indented from the main text. This format is not followed in braille; the extract is simply set off from the body of text by one blank line at the beginning and the end. Do this by using a skip-line indicator in place of a paragraph indicator at the beginning and the end of the extract.

This extract takes up more than half of print page 222. In the last paragraph of the extract, the text reads "...cut at 1 1/2 to 2 inches." For better braille, add a hyphen between the one and the half. This is discussed in detail in Section 8, Part 2.

Braille entry codes and translator controls

Move to the end of print page 224. Under the heading "For Sunny Areas" there is a list of ground covers. The name of each ground cover appears in boldface, followed by the Latin name in italics. Braille boldface and italics entry codes, beginning with the greater-than sign, are used here. Translator controls, beginning with an underbar, are used around the Latin words so that they are translated into grade 1 braille. Braille entry codes are explained in detail in Section 8, Part 3; translator controls are discussed in Section 17, Part 2.

Transcriber's note

The last item in our excerpt from Chapter 8 of The Wisconsin Garden Guide is a captioned picture. Move to the very end of chapter WGG SECT8 to see how a transcriber's note is used to explain this. Transcriber's notes are discussed in Section 13, Part 1.

For the final check, print WGG SECT8 to the screen and follow along with the print original. You have now completed the first step in the TranscriBEX process: entering and proofreading your data.

Part 4: MAKE$: Why and How

Throughout this manual, we say things like, "The \\runninghead command places the following text at the top of every page," and "The \\ssi command establishes indent to cell 5 and runover to cell 9." This is a good way to think of these commands. Technically speaking, however, the TranscriBEX \\ commands don't do anything at all. We designed them to be easy to remember. Option R - Replace characters, introduced in Learner Level Section 12, lets you find and change long li/s of things. When you use Replace characters with MAKE$ as the transformation chapter, it replaces the mnemonic \\ commands with intricate strings of BEX $$ commands. The BEX $$ commands do the actual work of formatting your text. To avoid unnecessary tedium, however, we continue to ascribe formatting power to the TranscriBEX \\ commands throughout this manual.

You now have your original print \\ command chapter, WGG SECT8 in our case, on disk. You need another freshly initialized disk for your intermediate $$ chapters. Label this disk "Intermediate $$ chapters" (or something similar). You can consider this a sort of "scratch disk" if you like. You must save your original and your fully translated chapters, but saving the intermediate $$ chapters is unnecessary.

Here's how to proceed: Your BEX disk is in drive 1; the disk with your corrected copy of WGG SECT8 is in drive 2. Choose option R - Replace characters on the Main menu. BEX reads that portion of the program into memory. When it's ready, you're prompted for a source chapter; in this case, answer WGG SECT8. BEX prompts again for a source chapter. Since you only have one source chapter, enter <CR> to the second "Drive or Chapter:" prompt. Then you're prompted for a target chapter. Specify that your target chapter is in drive 1 (that disk isn't really there yet, but it will be in time), and enter the same chapter name with a dollar sign added at the end. In this case, type 1WGG SECT8$.

Take your BEX disk out of drive 1, and insert your TranscriBEX disk. You're prompted for a transformation chapter. Here's where MAKE$ comes in. Precede the chapter name with the digit 1 so BEX looks on drive 1, and type 1MAKE$. This chapter is loaded into memory.

BEX now prompts, "Continue? Y." Take your TranscriBEX disk out of drive 1, and insert your scratch disk. Press <CR> to accept the continue prompt. You'll hear a series of clicks as your \\ commands are replaced with $$ commands. (If you're working on a 2c you may need to turn the volume up to hear this.) This process takes about six and a half minutes for chapter WGG SECT8, or about a minute per BEX page.

When Replace characters is finished, the Main menu prompt reappears on your screen. Take your scratch disk out of drive 1 and re-insert your BEX disk.

Part 5: Braille Translation

Now that all the \\ commands are transformed into $$ commands, it's time to do the braille translation. The Grade 2 translator recognizes all $$ commands, and leaves them alone. Only your text is translated. This is also where TranscriBEX handles the special braille translator controls and boldface entry codes that appear at the very end of WGG SECT8. They're described in Section 8.

Here's how to proceed: Choose option G - Grade 2 translator on the Main menu. When this segment of the program is loaded into the computer's memory, you're prompted for a source chapter. Take your BEX disk out of drive 1, and reinsert your scratch disk. Enter the name of the chapter you created in the last step, in this case 1WGG SECT8$. (Again, remember to specify that it's in drive 1.)

The next prompt is for a target chapter. Take your flippy disk with your original print chapter on it out of drive 2. Turn it over and reinsert it. If you're not using a flippy disk, insert a third freshly initialized disk at this point. (Label this disk or side "Grade 2 chapters" or something similar.) Enter the same name as the source chapter, adding a 2 at the end; in this case type WGG SECT8$#2. You'll see "Starting to translate" on the screen.

Sit back and relax for a few minutes. This process is fast, but not instantaneous. Chapter WGG SECT8$ is seven BEX pages. The translation takes less than five minutes in this case. How long it takes on your chapters is directly related to their size. The translation process is relatively silent; you hear some whirs coming from the disk drives, but none of the clicks you heard during Replace characters.

At this point, your print text has been turned into correctly formatted grade 2 braille--you think. But don't rev up the embosser just yet. You can make certain your text looks the way you want it to without an actual physical printout.

Part 6: The Braille Previewer--How To Use It

One of the best things about TranscriBEX is that it allows you to see what's going to appear on the braille page before the text is actually sent to the external brailling device. Think back to our discussion of screen braille and computer braille in Section 2. As we said there, when you send a braille chapter to a print device, it appears in screen braille. The braille previewer uses a print medium, so you don't see dots on the screen. Rather, you see an exact print representation of each cell. The format is exactly the same on screen as on paper because the print representation of braille is one-to-one: each of 64 possible braille cells is represented by one, and only one, print character. Screen braille looks a little strange when you first see it. But even experienced transcribers find that it is easier to proofread screen braille than dots on the screen. In screen braille, the cells representing contractions stand out clearly from the cells representing individual letters. In addition, the Video Mode braille previewer displays a "cheat sheet" which says what the ASCII characters represent in braille.

Using the braille previewer saves a lot of time and paper. This step is why we had you set up a braille previewer in Part 2. (You did set it up when we said to, didn't you?) As we mentioned there, you must have an 80-column card to use this feature of TranscriBEX.

Choose option P - Print on the Main menu. Enter the name of your formatted, translated chapter, WGG SECT8$#2 in this case, at the chapter prompt. When the prompt says, "Which printer?" answer Printer 3 (or the printer you defined as the braille previewer in your configuration). The screen braille version of your text appears on your monitor, formatted exactly as it will appear in the embossed version. Line 22 on your monitor displays the BEX page number. Use this as a guide for locating errors.

Braille files to braille previewer

The braille previewer is for use with braille files. Unless you have TranscriBEX commands in your text telling it otherwise, the formatter uses the standard braille paragraph values of indent to cell 3 and runover to cell 1. It doesn't skip a line between paragraphs. When your text contains formatting commands, the braille formatter uses them to format your text. It formats identically whether you Print to a real brailler or a braille previewer.

The Video Mode brailler displays your text one page at a time. However, the Apple monitor can only display 24 lines at a time. When your form length is 25 lines or longer, press the down arrow to see the rest of the page. Since the braille page number appears on the last line of the page in textbook format, it's important to look at the last line on at least the first couple of pages. When you're ready to look at the next page, press the spacebar.

Error message

When you enter a BEX $$ command incorrectly, the formatter simply executes what's there as best it can, and goes on. The results can be pretty strange! This is another reason we developed the TranscriBEX \\ command system. The consequences of entering a \\ command incorrectly are much less severe. When you print text containing an incorrect \\ command to the braille previewer, the message ******ERROR****** appears. This is very easy to pick out on your monitor. Note where it appears in your text (the BEX page number appears on line 22 of the display), and go back to your original \\ chapter to correct it in the Editor. Make any other corrections you need to, and send your chapter through Replace characters and the Grade 2 translator again.

Print your re-formatted, re-translated chapter to the braille previewer as many times as necessary. When it looks the way you want it to on the braille previewer, it's ready to be embossed. What you saw on the previewer in screen braille is exactly what you get on the actual brailling device in dots.

Print files to braille previewer

The first several times you use TranscriBEX, we suggest you go through all the steps in the order presented. There is a shortcut, but please don't use it until you feel you have a thorough grasp of what each step in this procedure does. The shortcut is this: You can print the formatted, UNtranslated chapter (WGG SECT8$ in this case) to the braille previewer. This gives you an approximation of how your text is formatted. All the formatting commands are in place, and are executed in this printing. However, because translation has not yet taken place, the number of characters in the chapter at this point is not the same number that will be there in the final edition.

You can see whether or not your format is essentially working. The ******ERROR****** message shows up at this point if you've entered \\ commands incorrectly. But you can't determine anything from this that requires knowing the exact number of characters that occupy a particular space. You can't determine where lines or pages end, for example, or how many braille lines a long table of contents entry occupies.

More than one brailler

You can define more than one previewer in a configuration. This is especially handy if, for example, some of your material is embossed on site, and some of it is sent elsewhere to be embossed. Let's say you're producing a newsletter in print and braille on site with an Image Writer and a Cranmer. You're also doing data entry for books to be embossed by National Braille Press.

Configure Printer 1 as a Generic printer and Printer 2 as a Cranmer Brailler, with a carriage width of 41 and a form length of 25. Configure Printer 3 as a Video Mode Brailler with carriage width of 41 and a form length of 25, the same as the Cranmer. Configure Printer 4 as a Video Mode Brailler with a carriage width of 38 and a form length of 28; these are the values used by National Braille Press.

When you want to preview your newsletter, send it to Printer 3. When you want to preview a book you're transcribing, send it to Printer 4. When it looks right on Printer 4, you can copy your disk and send it off to National Braille Press to be embossed, confident that you know EXACTLY how it will turn out.

Part 7: Emboss It

NOW you have correctly formatted, translated, wonderful, grade 2 braille. You don't just think so--you know it! Select option P - Print on the Main menu, and enter the name of the formatted, translated chapter you want embossed. In this case, enter WGG SECT8$#2. This time, specify the printer number of your actual device. If you used the configuration in our sample, this is Printer 2. Et voilá! Out comes the braille!

Detailed information on configuring and using every braille device we know about appears in the BEX Interface Guide.

Part 8: The End of Our Example, and Beyond

This Section has used a lot of space to explain the TranscriBEX process. The process isn't all that complicated, but accuracy is important. To sum up, here's a sample dialogue which contains all the steps.

Main Menu: E

Editor

Drive or Chapter: WGG SECT8 <CR>

There are 7 pages

Edit on page 1 <CR>

do data 5try and proofr14+

Main Menu: R

Replace characters

Drive or Chapter: WGG SECT8 <CR>

Drive or Chapter: <CR>

Target chapter name: 1wgg SECT8$ <CR>

remove BEX disk from drive 1 and 9sert TranscriBEX disk

Use transformation chapter: 1make$ <CR>

remove TranscriBEX disk from drive 1 and 9sert your scratch disk

Continue? Y <CR>

click click click

Chapter WGG SECT8 done

Replaced 72 times

remove your scratch disk from drive 1 and 9sert your BEX disk

Main Menu: G

Grade two translator

remove your BEX disk from drive 1 and re-9sert your scratch disk

Drive or Chapter: 1wgg SECT8$ <CR>

Drive or Chapter: <CR>

remove your pr9t data flippy disk from drive 2 and turn it over1 or put 9 another floppy disk

Target chapter name: WGG SECT8$2 <CR>

Target chapter name: <CR>

Starting to translate

pause--dr9k some coffee Main Menu: P

Print

Drive or Chapter: WGG SECT8$2 <CR>

Drive or Chapter: <CR>

Which printer: 3 <CR>

exam95 your chapter on the braille previewer2 make corrections on the orig91l chapter and s54 it his the hoops as many times as necessary

Main Menu: P

Print

Drive or Chapter: WGG SECT8$2 <CR>

Drive or Chapter: <CR>

Which printer: 2 <CR>

out comes the formatt$ grade 2 braille6

Operating on more than one chapter

Of course, the time will come when you would like to perform some of these steps on more than one chapter at a time. No problem! Instead of typing each chapter name when you're prompted for "Drive or Chapter," enter the disk drive number. BEX presents you with a numbered list of chapters. Enter the number of each chapter title you want to work on. There's one change in the dialogue when you've specified more than one chapter to be Replaced or translated. Instead of being prompted for "Target chapter name," you're prompted for "Target chapter naming method."

Herein lies one of the major virtues of establishing a consistent system for naming your intermediate $$ chapters and your translated grade 2 chapters. One target chapter naming method, I, is to Individually enter each target chapter name. When you're doing ten chapters, that's much too much work. Instead, use A, and Add a character to the end of all the chapter names in the list. When you Replace with MAKE$, the target chapter naming method is "A$"; when you translate, it's "A2".

Further details on target chapter naming methods appear in Learner Level Section 11 and User Level Section 4. A method of obtaining a list of only those chapter names ending with a particular character is also discussed in User Level Section 4. This is a very good argument for naming all your chapters of a particular type in the same way.

When you adhere faithfully to the chapter naming methods described in this Section, you never have to wonder whether or not a particular chapter has been through MAKE$ or been translated; the chapter name makes that obvious. Of course, following the recommended method of organizing types of chapters by disk makes your life easier still.

Another timesaver

You'll never go wrong following the steps in the TranscriBEX process exactly as described above. When you feel comfortable with all the steps in the procedure, you can save time and minimize disk juggling by copying MAKE$ onto your BEX program disk. Modify the process above accordingly.

WGG pages omitted

Section 8: Making Even Better Braille

Part 1: Overview

The braille system represents language in a medium that's tactile rather than visual. There is essentially a one-to-one correspondence between grade 1 braille cells, and inkprint letters and punctuation. Grade 2 braille is a refinement; some single cells represent letter combinations and whole words. Some combinations of braille cells represent something different than the sum of the single cells. For example, the braille number sign followed by the letter a doesn't mean "number a"; it means 1. Similarly, dot 6, dot 1 doesn't mean "comma a"; it means uppercase A. This code system pretty well covers the written English language.

However, there are a number of things which appear in print that, by their nature, are transmitted in the visual channel only. Colored type is an illustrative example. Blue print can't be made into blue braille. It can't be made into blue voice on the audio channel either, but the reader could say, "What I am about to read is in blue type; it's blue up to the point I tell you it isn't anymore."

The same basic strategy is employed in braille. Composition signs are used to tell the reader, "Hey! There's a distinguishing feature about this print. It continues until you come to the composition sign that says it ends." The assignment of composition signs to their various meanings is somewhat arbitrary, but they've been standardized. This frees the transcriber from having to invent ways to represent multiple typefaces, and frees the braille reader from wondering just what it is that the book is trying to show.

If you are doing direct braille data entry, read no further. This Section is all about how to get braille composition signs into your text that the translator doesn't place automatically. If you are doing direct braille data entry, you are entering the composition signs manually anyway. Turn directly to Section 17 for directions on direct braille data entry. From here on in, the text of this Section refers to print data entry only.

TranscriBEX's Grade 2 translator copes very well with translating most things that you type in at the keyboard. For instance, it automatically adds the uppercase composition sign (a dot 6 in braille, a comma in screen braille), to letters entered in uppercase on the keyboard. A digit entered on the keyboard is transformed to a letter preceded by the number composition sign. The number 9, for example, becomes dots 3-4-5-6, dots 2-4 (number sign, letter i in screen braille). TranscriBEX also has a set of commands to show braille italics, \\ib and \\if; their placement and effect are discussed in Section 6, Part 4.

There are some more complicated situations where the translator needs your help. We've designed a system of special symbols to allow you to enter braille composition signs directly into your text. This Section describes the various codes you enter in your text to help the translator deal with some very specialized inkprint situations, thus improving the quality of the braille transcription. The symbols you type into your text begin with the greater-than sign >. This signals the translator that something special is coming next. In most cases, the rest of the entry code is what actually appears in screen braille. Where it doesn't correspond with the screen braille, the rest of the entry code is mnemonic. For example, to place a braille check mark, enter >*. In this Section, we transcribe print data entry codes in computer braille enclosed in computer braille parentheses of with; all letters are lowercase. We enclose the braille results in braces {}. The material inside the braces is shown in print in screen braille; thus {#cm} appears in print as {number sign, c, m}.

In this Section we say a lot of things like: 'When "3m" is translated, it appears in screen braille as {#cm}; to force a letter sign, enter 3>;m; it then appears as {3;m}.’ This can be confusing, but there is a system here. When we talk about single print characters or short groups of them, we enclose the print in double quote marks: "". When we show you the exact symbols to enter in your text, we enclose them in parentheses: ( ). DON'T type the parentheses themselves, just what's inside them. When we show you what appears in screen braille, we enclose it in curly braces: {}. These braces don't actually appear on the braille previewer, just what's inside them. When we use spaces, in both the braces and the parentheses, they are significant. When you see a space in the parentheses, enter it in your text as part of that entry code. When no spaces appear in the parentheses, don't use them around the entry code unless you specifically want spaces to appear in your text.

When you are unsure what the translator is going to do with a particular chunk of text, don't forget your friend the Heading test! You can enter any short segment of text, with or without the special entry codes described in this Section, and see the result in screen braille instantly. Just be sure to include enough surrounding characters for an accurate translation.

Part 2: The Number Sign

The digits 1 through 0 are represented in braille by the lowercase letters a through j, preceded by the number composition sign. The braille number sign is dots 3-4-5-6. This appears in screen braille as the print number sign {}. When the number sign is used, it is in effect until turned off by a space, the letter composition sign, or a character which is never used as a digit. Characters never used as digits include lowercase letters k through z, all uppercase letters, and all punctuation. TranscriBEX automatically places the letter composition sign when it's needed after a number.

Some examples here should prove useful. "3" appears in screen braille as {#c}. "333" becomes {#ccc}. "3c" is correctly translated as {#c;c}. (The semi-colon is the letter sign in screen braille; further details appear in Part 6.) "3C" is translated as {#c,c}. The letter sign is omitted in this case because the capital composition sign makes it obvious that it's a letter; there's no such thing as a capital number. "3j" is translated as {#c;j}, but "3k" becomes {#ck}. The letter sign is omitted here because k is never used as a digit.

None of the examples above require you to manually enter any composition signs; the translator takes care of all those cases automatically. There are some cases, though, where one number is immediately followed by another, and the two numbers must be distinguished from each other. This is common in religious texts (such as the Bible) where two successive numbers are used to indicate chapter and verse. For example, "Paul 11:7" is read as "Paul, eleven, seven." The colon may not be used in this case in braille, nor can the numbers simply be run together. If you entered (117), it would be translated as {#aag}, and the braille reader would interpret it as "one hundred seventeen". This case calls for the number sign to be forced to appear a second time before the second number. Enter (11#7). Because the number sign is the same in print as in screen braille, no translation of the sign itself occurs, and the result is {#aa#g}.

When poetry or prose appears in print with line numbers in the margin, the line numbers must also appear in braille. The Code of Braille Textbook Formats and Techniques requires that the number composition sign be omitted in this case. The fact that the number is in a particular place is considered sufficient to tell the reader that it is a number rather than a letter. This is discussed further in Section 16, Part 4.

Part 3: Braille Italics

Sometimes, changes of typeface in print are only there to look nice; they don't mean anything. Setting the first word of every chapter in fancy Gothic script is an example of this. These embellishments are irrelevant to the braille reader, and should be ignored in transcribing. In other cases, however, changes of typeface are used in print to show emphasis. When this is the case, the same stress must be shown in braille. There are several ways to accentuate print, including underlining, boldface, and italics. When there is only one level of emphasis, use braille italics, regardless of how the emphasis is shown in print. (When distinction must be shown between two or more levels of emphasis, things get more complicated. See the Code for details on when, and Part 5, below, on how.)

When three or fewer words are to be italicized, the single italics indicator is placed before each word. The single italics indicator is dots 4-6. This appears as a period {.} in screen braille. When four or more words are to be italicized, the double italics indicator is placed before the first word and the single italics indicator before the last. The double italics indicator is two dots 4-6 in a row ...

The easiest way to get these symbols into your text is to use the \\ib and \\if commands, discussed in Section 6, Part 4. However, there are certain situations in which these commands won't work. They are inappropriate for italicizing only part of a word, for example. The alternative is to enter the italics indicators directly into your text.

The entry code for the single italics indicator is . (greater-than sign, period). The entry code for the double italics indicator is .. (greater-than sign, period, period). Use the single italics indicator to italicize a single word or part of one word. Place the italics entry code immediately before the first letter to be italicized. When the first italicized letter is not the first letter in the word, place a hyphen immediately before the italics indicator. Enter of-ar} (hyphen, greater-than sign, period) in your text. This becomes dots 3-6, 4-6 {-.}.

The double italics indicator is only used when four or more words are to be italicized, as described above. Use the single and double italics indicator entry codes whenever you find it more convenient than using \\ib and \\if.

The braille termination mark

When a whole word is italicized, the space at the end of the word tells the braille reader that the effect of the italics composition sign ends there. When the italics stop before the end of a word, you must insert the braille termination mark after the last italicized letter. The braille termination mark is dot 6, dot 3. It appears in screen braille as {,'} (comma, apostrophe). To place it in your text, enter >t (greater-than sign, lowercase t). Don't use spaces around the entry code unless you want spaces around it in your text. For example, to italicize only the middle syllable of "banana", enter ba->.nan>ta

Part 4: Foreign Words and Accented Letters

The treatment of foreign words and phrases in English text depends on whether or not the words are anglicized. Any word which can be found in the body of a comprehensive English dictionary is considered anglicized. Some examples are laissez-faire, señorita, and béte noire. Foreign words which are anglicized, or which are proper names, are contracted just as English words are. When an accent sign, tilde, circumflex or other diacritic mark is used in the word, use the generic accent sign in braille. The accent sign is dot 4; it appears immediately before the letter it affects. To enter the accent sign in your text, place an at-sign ` immediately before the accented letter.

Words which are not anglicized and are not proper names must be shown in uncontracted braille. Section 17 discusses how to do this. Whole texts or passages in a foreign language must be transcribed in grade 1 braille, using specific marks for each kind of accented letter. This is also discussed in detail in Section 17.

Part 5: Different Typefaces, Colored Type, and Shape Indicators

Braille boldface

When it is necessary to show boldface in braille, as distinct from braille italics, use the braille boldface indicator. There are single and double boldface indicators, which are placed just the same way as single and double italics indicators. When three or fewer words appear in boldface, use the single boldface indicator before each word. When four or more words are in boldface, use the double boldface indicator before the first word and the single boldface indicator before the last.

The single boldface indicator is dots 4-5-6, dots 4-6. It appears in screen braille as {_.} (underbar, period). Enter >_ (greater-than sign, underbar, period) where you want it to appear in your text. The double boldface indicator is dots 4-5-6, 4-6, 4-6. It appears in screen braille as {_..} (underbar, period, period). Enter >_.. (greater-than sign, underbar, period, period) in your text.

When you need to use a boldface entry code in the middle of a word, use the same method as for mid-word italics. Enter of-arwith (hyphen, greater-than sign, underbar, period). It appears in your text as dots 3-6, 4-5-6, 4-6 {-_.}.

Colored type indicators

Many textbooks use multi-colored type. It is not necessary to show this in braille unless the reader gains some sort of information from it. For example, when the text talks about "the words written in red," the braille reader needs to know which words they are. But when the first word of every chapter is printed in green, this is decorative rather than informative, and irrelevant to the braille reader.

When it is necessary to show colored type in braille, go through the book or passage and make a note of all the colors that appear. Use the first letter of the color as its code; for example, r for red, b for blue, and g for green. When this is not distinctive enough, use more than one letter. For example, if you must distinguish between blue, black, and brown print, use bl for blue, bk for black, and br for brown. There is no universal code of initial letters of color; the important thing is that you explain what you are doing in a transcriber's note and then follow it consistently.

The colored typeface indicator is dots 4-5-6, followed by the initial letter or letters of color; it is set off by spaces on either side. This appears in screen braille as space, underbar, letter, space { _x }--x stands for any letter here. The braille composition sign for a letter does not appear. Enter of >_l (space, greater-than sign, underbar, letter, space) in your text, just before the first word in colored type. At the end of the colored type, enter a space, the termination mark, space >t . The spaces must appear; they are not optional in this case.

Shape indicators

A similar initial-letter code system is used when geometric shapes are used in print to indicate something significant. The shape indicator is dots 1-2-4-6, followed by the first letter of the shape. This appears in screen braille as space, dollar-sign, letter, space { $x }. Again, the letter composition sign does not appear. Enter >$x (space, greater-than sign, dollar sign, letter, space) in your text. Again, explain the code system you are using in a transcriber's note. In general, use single letters, such as s for square and c for circle. When you must distinguish between a triangle and a trapezoid, use more than one letter. The important things are a good explanation and consistency, rather than a universal code of shapes.

Special heading typeface indicators

Sometimes headings are shown in different typefaces that are significant to the reader. The Code defines three levels of special heading indicators. The first is dots 1-2-4-6, 1-2-3. Enter (>h1) (greater-than sign, lowercase h, one) in your text. This appears in screen braille as dollar sign, lowercase l ( $l }. (Don't worry; the translator won't confuse this with the new-line indicator unless you enter the wrong thing. Remember, YOU ENTER what's in the parentheses; YOU SEE what's in the curly braces.) The second special heading indicator is dots 1-2-4-6, 1-3. Enter >h2 (greater-than sign, lowercase h, two). The screen braille is dollar sign, lowercase k {$k}. The third special heading indicator is dots 1-2-4-6, 1-2. Enter >h3 (greater-than sign, lowercase h, three). This is dollar sign, lowercase b {$b} in screen braille.

When you use a special heading indicator, place it between the heading command (\\hd, \\c, or \\mh) and the text of the heading. Use one space between the command and the special heading indicator, and one space between the indicator and the text of the heading. Use a termination mark >t between the text of the heading and the paragraph or skip-line indicator after it. Use a single space between the text and the terminator, and another between the terminator and the indicator. For example:

$s \\hdarheaThis heading is in Big Ball typefaceart$so

When you use the termination sign, colored typeface indicators, shape indicators, and special heading typeface indicators in your text, you must list them on a page of special braille symbols in the preliminary pages of the book. A transcriber's note should then explain the use of each indicator just before it appears for the first time in the text.

Part 6: Controlling the Letter Sign

In grade two braille, many isolated letters represent a word. For example, a lowercase n by itself stands for the word "not". When the text must show "really just the letter n," a letter composition sign, dots 5-6, is placed before the letter. This appears as a semicolon {;} in screen braille. Some additional uses of the letter sign include differentiating a letter which appears after a number (see Part 2), and before lowercase roman numerals. TranscriBEX's braille translator correctly places the letter sign in the vast majority of cases. When you need greater control of the letter sign, you can manually override the translator.

TranscriBEX's translator places the letter composition sign before all stand-alone (i.e. with a space on both sides) letters entered at the keyboard, EXCEPT a, i, and o. To force a letter sign to appear before one of these letters, enter greater-than sign, semicolon >; before the small or capital letter to which the sign refers. You can use this entry code whenever you want a letter sign but don't think the translator will provide it. When you use this code and the translator would have provided the letter sign anyway, it has no effect; you don't get two letter signs. Examples of other places you may find it useful to force the letter sign are to put one before a single letter followed by punctuation, or to place a letter sign before a lowercase roman numeral. (TranscriBEX automatically places the letter sign before roman numeral page numbers.)

To suppress the letter sign, enter less-than sign, semicolon <; just before the letter you want to affect. This is the only place you use the less-than, rather than the greater-than, sign. This is because it is the only place you are suppressing, rather than forcing, a composition sign. Again, this has no effect when TranscriBEX would have left out the letter sign anyway. Use it where you have noticed that the translator is more generous with the letter sign than you would like it to be. An example of where you may need to suppress the letter sign is before a single-digit uppercase roman numeral.

Part 7: Miscellaneous Signs and Symbols

There are some print symbols which may not be entered directly at the keyboard because the braille translation of that symbol has a different meaning. For example, the plus sign + is the same as the braille ing contraction. Some print symbols must be spelled out, such as plus, minus, and equals. Rule VIII of English Braille contains a list of approved abbreviations for weights, currency, and some other special symbols.

Entering most units of weights and measures as they appear in print yields quite readable braille. The meaning of "5 ft. 2 in." for example, is quite unambiguous. The Library of Congress standard, however, requires that this be written as "ft5 in2." The following is a list of symbols supported by TranscriBEX to generate braille which is as close as possible to perfect.

Complex numbers and mathematical symbols

Fractions: Use a slash between the numerator and denominator. Enter the fraction one-fourth as 1/4; enter ten one-hundredths as 10/100. For a mixed number (one which contains a whole number followed by a fraction), use a hyphen between the whole number and the fraction. Enter two and one-half as 2-1/2; enter eighty-five and five-sixteenths as 85-5/16.

Math operators: The common mathematical operators (+ - =) must be written in words, such as plus, minus, equals, times, and divided by. For example, the equation (2+5=7) must be written as "2 plus 5 equals 7".

Percent sign: The percent sign may be entered directly at the keyboard (it's over the 5). However, the percent sign appears before the number in braille, so enter Here % means the special print symbol for per cent.

Roman numerals: Uppercase roman numerals do not require a number sign before them. Enter them exactly as they appear in print. To suppress the letter sign before a single-digit uppercase roman numeral, use <; as described in Part 6. A letter sign is required before lowercase roman numerals. TranscriBEX places the letter sign automatically before single-digit roman numerals in text (and all roman numeral page numbers). To force a letter sign before a multiple-digit lowercase roman numeral, use >; as described in Part 6.

Weights, measures, and currency

Weights and measures: In general, the abbreviation is placed before the number, unspaced and unpunctuated. Enter "5 ft. 2in." as ft5 in2. For a list of approved abbreviations, see Rule VIII in English Braille.

Non-decimal British currency: Use the abbreviation for the largest stated value before the first number, and a number sign before each subsequent unit. For example, six pounds, eight shillings, ten pence is abbreviated in print as "L6 8s 10d". For braille, enter l16#8#10. To write "2s 2d", enter s2#2.

The braille apostrophe and single quote marks

Typewriter-style print uses one symbol to represent three things: the apostrophe, and both the open and close single quote marks. Braille uses a separate representation for each of these three symbols. The apostrophe is dot 3; it appears as an apostrophe {'} in screen braille. The open single quote is dot 6, dots 2-3-6, or {'} (comma, eight) in screen braille. The close single quote is dots 3-5-6, dot 3, or {'} (zero, apostrophe) in screen braille.

When you are doing print data entry, you use the apostrophe to enter all three. During translation, BEX's Grade 2 translator uses context to "guess" which symbol the print apostrophe represents. It guesses right in the vast majority of cases. An apostrophe between two letters is always interpreted as such; you never need to help it along. An apostrophe at the end of a word ending in s is also always interpreted as an apostrophe.

Some text, particularly poetry and dialectical language, uses apostrophes at the beginnings and ends of words. The translator is more likely to interpret these apostrophes preceded or followed by spaces as single quote marks. Whenever you want an apostrophe and you think the translator might use a single quote instead, use a greater-than sign just before the apostrophe, like this: ofar'with. The translator then knows exactly what you want. When the translator would have provided an apostrophe anyway, this has no effect; you never get two apostrophes.

Here's a sample of all the possibilities:

'twas wha' Boris' children feared, e'en worse than 'water-torture'.

The translator interprets the first two apostrophes as open and close quotes, so you enter:

>'twas wha>' Boris' children...

The translator copes correctly with the rest of the sentence.

Other symbols

Section mark: Print sometimes uses a fancy sort of S as a section marker. When it is necessary to show this in braille, enter >sec (the greater-than sign followed by three lowercase letters) in your text. Place spaces around the entry code as they appear in the print. This entry code becomes dots 2-3-4, dot 3, or lowercase s, apostrophe {s'} in screen braille.

Ditto mark: To show the print ditto mark two strokes beside each other resembling a quotation mark in print, enter >dit (space, greater-than sign, three lowercase letters, space) in your text. It becomes space, dot 5, dot 2, space. This appears in screen braille as {"1}.

Check mark: To show a check mark in braille, enter >chk (space, greater-than sign, three lowercase letters, space). In braille, this becomes space, dots 2-6, 2-6, space. This appears as {55} in screen braille.

Insertion mark: When the print uses a caret ^ as an insertion mark, do the same in braille. (The caret is above the 6.) To show an insertion, enter ^ (space, caret, space) before the inserted text. Enter a termination mark with spaces on each side >t after the inserted text. The print caret translates into dots 3-5, dots 2-6, the braille caret. It appears in screen braille as {95}.

Reference indicators

General reference indicator: Sometimes it is necessary to show a footnote or other reference mark where no special symbol appears in print. The braille symbol for this is two dropped g's in a row (dots 2-3-5-6, 2-3-5-6). It appears as {77} in screen braille. Enter >gr (greater-than sign, two lowercase letters) in your text. Place it after the referenced text (and any accompanying punctuation) and at the beginning of the referencing note.

Asterisks and superscripts: When the print uses an asterisk as a reference indicator, do the same in braille. Place the asterisk after the referenced word and any accompanying punctuation. When the print uses a number as a reference indicator, use an asterisk followed by that number for the braille. Don't leave a space between the asterisk and the number. Enter the asterisk * directly into your text from the keyboard. It becomes dots 3-5, 3-5 in braille, and appears as {99} in screen braille. In general, it should be placed after the referenced material and at the start of the referencing note. See the Code or English Braille for the specific rules of placement.

Part 8: FINETUNE Braille Translation

In Section 7, we outlined the steps in the TranscriBEX process: enter and proofread your print data, Replace your \\ commands with MAKE$, translate into Grade 2 braille, and preview. These steps combine to make very good, readable braille. In this Section, we've given you the means to tweak up your braille output one notch further by entering some special braille codes into your text. To get one degree closer still to perfect braille, there's another step to perform in the TranscriBEX process before embossing.

BEX's Grade 2 translator is quite good, but it's not perfect. To understand why, you need to know a little bit about how it works. When you choose options G or H on the Main menu, BEX loads two files: one is the translator program itself, and the other is the translator's tables. The tables are contained in the BEX chapter named ZQFOR. They help the translator program deal with the many subtleties of choosing appropriate contractions. But they're limited in size. We can not make a table entry for every possible word. If you're familiar with computer programming and feel adventuresome, you can attempt to modify the translator tables--more on this in Master Level Section 10, Part 2. (By the way, we always appreciate hearing in writing about any translator problems you encounter.)

Modifying the translator table is really a job for a programmer. But regular computer users have a tool that's better-fitted to their hands: Replace characters. The purpose of the FINETUNE transformation chapter is to find and fix mis-translated words in braille chapters. The FINETUNE chapter supplied on your TranscriBEX disk improves the braille translation of 21 words that we've noticed need fixing, but which are obscure enough that we don't want to reprogram the translator table to include them.

Using FINETUNE

When you want to improve the quality of braille translation, you use Replace characters with FINETUNE as your transformation chapter. Your source chapters are your final, formatted grade 2 braille files, the ones with names like SAMPLE$2. Usually, it's a good idea to use different source and target chapters when Replacing. For example, if you didn't keep your original \\ command chapters when you used MAKE$, you would have a very hard time making corrections later. But in this case, Replace characters is making corrections on your grade 2 chapters for you; it's not substantially altering the nature of your entire chapter. Because you're improving your text, you don't need to keep your source chapters. When you're Replacing just one chapter, type the same name at the "Target chapter:" prompt. Here's a sample dialog; you begin with your BEX disk in drive 1 and your data disk in drive 2.

• Main Menu: R
• Replace characters
• Drive or chapter: SAMPLE$2 <CR>
• Drive or chapter: <CR>
• Target chapter: SAMPLE$2 <CR>
• [Remove your BEX disk and insert your TranscriBEX disk in drive 1 at this point]
• Use transformation chapter: 169netune <CR>
• Continue? arY <CR>
• Starting to replace...

When you're Replacing more than one chapter, BEX prompts for a "Target chapter naming method" instead of a "Target chapter." Answer S for Same name for source and target chapters.

Inside FINETUNE

Section 8 of the User Level discusses the structure of transformation chapters in detail. To summarize, a transformation chapter contains some number of transformation rules. Each transformation rule consists of a "Find" string and a "Change to" string. For each transformation chapter, you must reserve one character to use as a "terminator." Each transformation rule contains two instances of this terminator: the first defines the end of the Find string and the second defines the end of the Change to string. Each transformation chapter contains two "extra" terminators. One must be the very first character: it tells BEX which character is reserved as the terminator in this particular transformation chapter. The very last character must also be the terminator: it signals the end of the list of rules. In FINETUNE, the terminator is <CR>.

Make a copy of FINETUNE, and examine it in the Editor. The first character is always the terminator, in this case, <CR>. The first rule finds SEATO and changes it to SEATO . It's wrong to use the ea sign for the abbreviation of the Southeast Atlantic Treaty Organization because it's an acronym. The second rule is a similar refinement: because "radar" is also an acronym it expands the ar sign to the two letters a and r. Since the first rule's find string begins with the double caps sign (two commas) it only changes SEATO when it's all caps. When dealing with a braille chapter, it's easier to write one rule to handle varying capitalization than it is with print, since the caps signs precede the word itself. The second rule begins with the letter r. This one rule matches RADAR, radar-controlled, and RadarRange with equal ease.

Sometimes, the translator encounters a series of three letters in which either the first and second letters or the second and third letters could be contracted. It has to make a choice; it doesn't always choose correctly. There is a series of words in the FINETUNE chapter which begin with the letter combinations bea, bed, ben, and ber. The translator must choose between using the be contraction, or ea, ed, en, or er. Most of the time, it copes; the list in FINETUNE represents the cases we've discovered that confuse it.

Adding to FINETUNE

When you discover that the Grade 2 translator is mis-translating something, add an entry to the FINETUNE chapter to fix it. Please don't change the FINETUNE chapter unless you're sure that what the translator's doing is really wrong. As with any subtle format issues, it's best to rely on the word of a skilled transcriber or Library of Congress-certified proofreader. Once you know you want to make an entry, actually adding to FINETUNE is quite straightforward.

Make a copy of FINETUNE on a separate disk. Before you start changing it, prepare a list of the errors you've found and their replacements. If you're comfortable with the six-key braille keyboard, then change to braille keyboard mode and braille screen mode for entering changes. (Details on braille entry appear in Section 17). If you're not familiar with braille entry, then prepare your list with screen braille in both "Find" and "Change to" sequences. Use the "Screen Braille Chart" on the TranscriBEX Reference Card when you need to translate dot-by-dot information given to you by a proofreader into screen braille equivalents.

Here's a step-by-step sample of adding one transformation rule to FINETUNE. While transcribing an Italian cookbook, you discover that the Grade 2 translator mis-translates "linguine." English Braille forbids the use of the ing sign across a syllable boundary, but the translator creates linguine.

Edit the copy of FINETUNE. Move your cursor to the very end with control-A spacebar. The last two characters are <CR>s. Place your cursor on top of the last <CR> and type: linguine followed was<CR>. You've just entered the Find string. Now type the Change to string: linguine followed was<CR>. The crucial final step is typing one more <CR>. For FINETUNE to work corrrectly, it must end with two <CR>s.

Quit the Editor, and test the modified FINETUNE chapter. Use Replace characters on one of your final, formatted braille cookbook chapters, but use a different name for source and target chapter. You don't want to overwrite your almost-done COOKBOOK$2 chapter until you're sure FINETUNE is working right. Use a name like COOKBOOK$26 to make this target chapter distinct. Carefully proofread the FINETUNEd chapter to make sure that you haven't unintentionally created a nightmare. When you're satisfied with the changes you've made, copy the modified FINETUNE on top of the previous FINETUNE on your TranscriBEX disk.

A word of caution

Any time you write a transformation rule, it's important to give enough surrounding text so that you don't unintentionally change more than you want to. In the course of writing this manual, I got sick and tired of typing "print page indicator," so I wrote a rule to change "ppi" to "print page indicator." I rolled merrily along until I wrote a chapter in which I talked about "flippies" and "floppies." Of course, when I ran it through Replace characters with my transformation chapter, I suddenly had a chapter full of "fliprint page indicatores" and "floprint page indicatores"! I should have used a space before "ppi" in my transformation rule.

This is especially crucial in FINETUNE. For example, if you tried to shortcut changing "linguine" to "linguine" by just changing "arfuling" to "ling", you'd also inadvertently change "filling", "fling", "billing", "mudslinger", "willing", and who knows what else! Be sure, and BE SPECIFIC!