One result is that a reader can customize the scroll “on the fly” for ease of reading. Magazine page columns are set up so that width and type size produce line lengths that most will find comfortable for reading, with a next line’s beginning easily moved to from a finished line’s end. You can make a web-scroll’s width, given your chosen type-size, comfortable by grabbing the browser’s or frame’s right hand edge and dragging it left or right. A long skinny Browser on your desktop? Sure, why not? And it won’t effect the printing, which is done on cut sheets and independently of the screened scroll.
This new millennium thinking is the context for thinking about and writing with eWriter — because it is a 21st century eTypewriter or a textwriter, htmlwriter, and xmlwriter.
Danger: Surrounding selected text with tags.
Q. I find that I can use some HTML tags from the menu and have selected text surrounded, but it doesn’t work for all of them. If I hit Form, it will replace the selected text with the tags. And Paragraph gets a warning dialog. Where is a list? And that replacing of the selected text is dangerous.
A. As of 0.9l+, the version of eWriter that is available as I write this, there is only this paragraph in the HTML Help topic (WinHelp, F1).
Italics and Bold put you between the tags with an I or B in them. With these two, however, you can do plug and play. If you select text first, the tags will embrace the text that was selected. Headings and Comments can also wrap selected text. Row frame, header cell, and row cell wrap selected text, as will List item tags.
You can see that I just kept adding in new ones. To make a list of it, it’s
The danger is real. Other HTML menu items will replace selected text. And UNDO won’t work because the typing and formatting is complex. For HTML Block or FORM (block), only the making of an empty line between tags is “undone.” I will fix this in 0.9m so that nothing happens and a dialog tells you it is because you have selected text. For now, Beware!
Ctrl+P or HTML/Paragraph is another matter. It would be very easy to become confused and hit Ctrl+P while something is selected. Actually, at the top of the XML menu, where some HTML tricks are in a section of the menu, you have items that will surround selected text with tags and allow you to use either <P>...<P> or <P>...</P>. That’s why I had the dialog pop up for this one and not others.
All Tagset menu items that you add will handle selected text intelligently. that is, if you have a tag pair as the item, selected text will be surrounded. If you have either an HTML empty or an XML empty, the highlighting will be turned off and the tag will be at the end of the text that was selected.
I've corrected this for all items on the HTML menu. If they don't surround selected text, they will turn off selection and simply push the text that was selected. The correction is in 0.9m which went up at TUCOWS and Coriolis on November 3 and will go up at 4D Solutions on November 10.
On Dec. 17, I've added this surrounding of selected text to the pairs of parentheses, brackets, curlies, quotes, commas, and asterisks. This and other features and fixes are for 0.9o to be posted in early January. I will email an 0.9n+ on request.
A built-in Table Editor — one click and two numbers.
Q. Since eWriter is both a text editor and a word processor, I wondered about the table editor that all major word processors have. I know that the HTML menu has all the tags for making a table, but that is hardly enough.
A. WinWord, PageMaker and other word processors and publishers do have table editors and these are difficult if not buggy, always being upgraded even apart from the major product upgrades. eWriter does have a table maker.
It’s in the upper section of the XML menu. It’s titled “n x m Table Frame” and when you click it two input boxes come up in sequence. Supply number of rows and number of cells and eWriter types out the table frame as if it were a punctuation mark. You go on typing. When you come back to fill in data, you find that each row and cell has a comment behind its start tag giving its number so you can find your way around easily.
You will have to put in your attributes, of course. If you build up a table of n x m rows and cells, fill in the attributes you want, perhaps header cells, and are going to reuse the table, ...put it between comment tags below the </HTML> tag or in another file. Then, you can paste it in anywhere as a mini-template.
My wife is a librarian at a community college and has built the library’s website (in eWriter) because nobody else was doing it. She’s currently working on the Periodicals List. Here is an often reused and filled table.
| Title | Holdings | Locations |
|---|---|---|
| BILLBOARD | 1 year + current | LRC 4 |
| BIOSCIENCE | 1969- |
LRC 4: 5 years + current LRC 3, Microfilm: 1969- |
| BLACK COLLEGIAN | May 1975- | LRC 4 |
| BLACK ENTERPRISE | 1975- | LRC 4 |
| BLACK ISSUES IN HIGHER EDUCATION | 1991- | LRC 4, Prof. Lib. |
| BLACK SCHOLAR | November 1969- |
LRC 4: 5 years + current LRC 3, Microfilm: Nov. 1969- |
| BULLETIN OF THE ATOMIC SCIENTISTS | 1971- |
LRC 4: 5 years + current LRC 3, Microfilm: 1971-1993 |
| BUSINESS WEEK | 1970- |
LRC 4: 5 years + current LRC 3, Microfilm: 1970- |
| BYTE | 1992- | LRC 4 |
Spell check ...and fancy printing.
Q. What about spelling checks? I know one of your eManual’s HTML pages deals with that, but there doesn’t seem to be anything built in.
A. You’re right, eWriter doesn’t have built in spelling checks or, for that matter, fancy printing. It has a wired in word counter, but that’s only useful if you’re not using tags, as it doesn’t skip them. The spelling and printing warrant a little discussion.
There is a very fancy print dump, of course. No matter what type size you use on screen or if you use color reversal, the print will be black on white and 10 point. Well, that’s black print on whatever color your paper is. You’ll get page numbers upper left if there is more than one. And a file name up there, too. If eWriter comes to a line that is over 84 characters, you are given a choice. Break and push the line or have a substitute line calling it a skip and telling the length. You don’t want to push the header in a Winword produced RTF file. This gives you a fast, good print of your “engine” page or HTML/XML “manuscript,” and you’ll print the "interface" copies from your (web) browser. If you want to do fancier, word processsor type printing of your “plain text,” you can set up anything, Write, WordPad, Winword, Ami Pro, whatever, as a hot key tool on your Tool menu, Copy All to the clipboard, fire up that tool, paste, and make your interface copy.
Meanwhile, eWriter stays small (for a Windows 95 program) and doesn’t cart along DLLs, OCXs, or wagonloads of other baggage.
Spelling... is handled in much the same way. I keep Winword 7.0 (before the 9? bloatware packages) as a tool on my Tools menu and the F7 hot key. 7.0 was the first 32-bit Winword, and it has that underlining in red of questionable spellings. I do my Copy All, fire up Winword, paste, and all the spellings to check are underlined. I do not correct them in Winword then and try exporting plain text. You don’t really know what will be exported or how.
I check the word, using Winword’s check if I need to, Then I AltTab back to eWriter, Shift+F3 to find the bad spell or typo, correct it, save with F2, and Alt Tab to Winword. It’s about as quick as using a dialog, and I don’t cart along dictionaries and (probably) a DLL or OCX to hold a speller.
If you want to simply move the text to something that will let you run the internal spell-check and then export your file back to you, use Win 95’s FAX Compose. It’s plain text. Thing is, it may use Winword’s speller and so may not do this if you haven’t Winword. I’m not sure about any machine but my own. Until fairly recently I used an old eighties IBM WordProof that’s smaller, faster, and in every way better than anything knewer I know of and had 129,000 word dictionary. Winword’s underlining got me. And I had to make a batch file that would move WP, dictionaries and all to where my file was and then delete them ...because the path for files going in had to be short.
I’m serious about my typewriter model in software. It’s just that, now, the extra things beside a typewriter on a desk like a bottle of white-out or rub-on letterset copy and such c’n be replaced by, say, a Winword (Word for Windows). My typewriter is for a writer, too, not a secretary or a clerk typists. Nothing wrong with anybody doing any of those jobs, but a writer wants to “arrange” his tools, including extra things, like the pan of hot water he soaks his feet in while writing --as somebody was said to do. With eWriter, you’ve got to “sharpen your pencils,” as it were. Get things set up to work.
Q. I know that eWriter is for writers and not, say, data entry people, but sometimes data entry typing will have to be done. And that is especially true for a lot of uses of XML (SGML-lite) for packing data into documents.
A. Say that you have a nested set of elements, making up an object model or a tree. You’re going to “fill in” a series of these with the “data.” Your best bet is to type up the object (arranged elements) once and keep it in a crib sheet file. Then, move it to the clipboard and paste a stack into your document. Start at the top, and fill in all the data, working your way down the stack. A trick to speed this up is to use some symbol at each fill point such as the @ or * or % or | that won’t be “in” your data. Then, use Shift+F3 once and F3 after that to go from fill point to fill point. The Find selects the symbol and if you’re bringing the copy from some other file via the clipboard, the paste will replace the symbol.
Every situation will call for its own design, so you have to approach even data entry with a writer’s eye and mind.
Inserting data in “free form” documents.
Q. In XML documents, I want to put into a free form memo or letter an address that will involve a complex nest of tags so that a utility that extracts addresses from the memo-base can insert the pieces in the fields of a record in an ordinary database. I know that I put my tags on the Tagset menu, but my question is ...how do I write this “nest” into my free form memo text?
A. Here is where thinking of tagging as punctuation marks, of tag-pairs as frame punctuation, and eWriter as a typewriter will really free you up to write naturally on two levels. In the Nov./Dec. Visual Developer Magazine, Jeff Duntemann writes in a sidebar Why Programmers Should care About XML, “The point of XML is to allow document authors to build a layer ‘behind’ the human meaning of the document that says, ‘Heads up, bitbanger; this is a street address.’”
That “layer” is the tagging that chunks the text but doesn’t affect the display of the text unless the tags are associated to formatting styles.
I just sent the original to <ADDR><NAME><FIRST>Jeff </FIRST> <LAST>Duntemann</LAST></NAME> at <STREET>14455 North Hayden Road, Suite 220</STREET>, <CITY>Scottsdale</CITY>, <STATE- SHORT>AZ </STATE-SHORT><ZIP>85260-6949</ZIP></ADDR> and he should have it by Monday. [This "text" will be seen in a browser that doesn’t "see" the tags --gf]
This will be seen as:
I just sent the original toand he should have it by Monday. [This "text" will be seen in a browser that doesn't "see" the tags --gf] at Jeff Duntemann 14455 North Hayden Road, Suite 220 ,Scottsdale ,AZ 85260-6949
Name tags carefully. I first did the above example using <ADDRESS> as my nest-root. The HTML browsers, of course, recognized that. My address was perfectly rendered (except for being italicized) but as a single piece, which made the line break after “original to”.
I set my eWriter up to type that example. I put in the ADDR, NAME, FIRST, LAST, STREET, CITY, STATE-SHORT, and ZIP tag-pairs. I had a couple tags already on there, so these became 3, 4, 5, 6, 7, and 8.
I typed Alt+G, 3 to get the ADDR pair. I immediately typed Alt+G, 4 and Alt+G 5 to get the NAME and FIRST tags nested. I typed Jeff and skipped behind </FIRST> and typed Alt+G, 6 to get LAST. I then typed Duntemann and skipped past </LAST>, and so proceded on until I skipped past </ADDR> and the “nest” was complete.
When typing in the amalgam of the ordinary layer of text’s meaning and that under-layer, a higher degree of “mindfulness” is required, but a writer should be pretty much aware in such “layers” of meanings behind meanings, anyway. Reading isn’t just puzzling out a sequence of words ...and writing can’t be any such thing. The Typing of Alt+G, # becomes as much second nature as typing Shift+/ to get a ?. And seeing a frame of two tags appear around your cursor (caret) where you will type the next phrase or word or datum will soon seem as natural as seeing that cursor in the first place.
When typing in nested tag-pairs, it is convenient if you can skip the innermost End tag with a single key stroke, just as you can go outside a single tag-pair by using the End key. I've added just such a function to both the Edit and Tagset menus. Ctrl+Period (Ctrl+.) is the key. This will be in 0.9n, but if you have 0.9m, which went up at most locations on Nov. 3 or the 0.9l+ at 4D Solutions until Nov. 10, but want this key, email me and I can email you a zipped copy of ewriter.exe only ...with the skip key.
I’ve said from before eWriter’s birth (as early as in the documentation for PocketPad 1.1) that in the coming century we are going to “read” and “write” very differently — and, now, XML is entering the picture. Stay light on your feet. Multiple paradigm shifts are the rumblings you feel under them.
Q. When eWriter offers to color the HTML/XML tags, it warns that the process can be time consuming. And it does this only on loading, so that newly typed tags aren’t blue.
A. It’s slow because every character in the file must be read and checked to see if it is a “<” or a “>”. Characters between these are turned blue. This is done inside Microsoft’s RichEdit control after the loading of the file is done.
It might be faster to use a filter on the loading file, except that I’d have to write in RTF code and that, too, could be time consuming.
This is done only on loading. It’s for convenience in proofing. Tags are punctuation, and to see them go blue as you write would be distracting. It would slow typing down, too, and, then, you could never type either key in your text.
Believe me, under that slow tag coloring on loading there is a seething tangle of snakes to sort out. You won’t often have an .htm, .html, or .xml file that’s long enough to take more than a minute to finish. On that rare occasion you can go for a cup of coffee.
You can stop the highlighting by hitting the ESC key. Or you can watch the progress by clicking on the white area (which is a panel that vanishes) and watch the tags being colored. The cursor returns to the top of the file when it’s done.
A writer sharpens his or her pencils.
Q. I know that if I am working on a batch of e-mail memos or other documents, I can have them all in editors and move among them to write or to copy data or text from one to another. But how can I use eWriter’s capabilities to help me with the more complex writing of XML documents?
A. A writer has always made something of a ritual of sharpening his or her pencils, or maybe quills, and getting paper, or parchment, ready, other tools. And while doing these focusing activities to prepare to write, the writer is organizing materials in his or her head, too.
You have quite a bit to do to prepare to work on an XML document or a group of them. You must put in your core Tagset, of course, in the Tagset menu. At first, you should build such a set so you understand just how the inner syntax works. But then your pencil sharpening will be to look at how the tagset is stored in ewriter.ini and write sets to that format in a file, swapping them in and out of the .ini file. Whether building your set or swapping it in, you have to have the menu (and keys) to work from.
You’ll also set in your main boilerplate cylinder to be the “usually active” set of eight boilerplates. Then, in the first several editors (in front of the documents you work on), you’ll want DTDs and, behind them, crib sheets. In fact, this group of DTDs and crib sheets can be a saved desktop group (see Window menu). This “desktop” group is in a file in eWriter’s directory with a .dsk ending. Give the file a name suggesting (or being) the document type. Pencil sharpening, then, involves pulling in this .dsk file (Restore Desktop) and loading your document file or files behind it. You may have other “reference” files in that group, too. Model documents or documents you may take text or data, even whole “nests”, out of. In short, you build up a working and thinking and imagining environment.
All the while, you’re mulling over what you will write in your innards. You may find that, after this sort of preparation, you write more quickly than you have before, but with a “sure footedness” that gets you down the trail with precise steppings. You may not match a Mozart writing out a composition without erasures or a Frank Lloyd Wright making all the drawings of the Kaufman house in under three hours (after three months of thinking about the ground, every rock and tree), ...but you should find a new quality of sharpness and quickness in your work.
What’s an e-letter ...and why?
Q. I know that many, even the oldest, papers in the eManual and elsewhere talk about the e-letter. But what is an e-letter and how is it something other than just e-mail?
A. In a purely mechanical sense, an e-letter is a letter sent by e-mail or, in printed form, by U.S. Mule, after being printed from a browser, that has embedded HTML tags as framework and punctuation. It’s intended to be read both in a mailer’s (or other) text editor and in a browser — where it takes on the “word processor” look. That involves sorting out tags and content and enabling a reader using the text editor to keep the tags in peripheral vision, acting like punctuation.
My own habit, is to not build up the framework as I might do ordinarily using the “format pad” that’s on the Number Pad using the Ctrl key. That build up is very fast. Ctrl+NP7, Ctrl+NP9, Ctrl+NP8, type title, drop below the </HEAD>, Ctrl+NP3, type the body of the HTML document. I use a couple boilerplate entries for e-letters to print these lines with, say Ctrl+Top3, Ctl+Top4:
<HTML><HEAD></HEAD><BODY BGCOLOR="FFFFFF"><BR></BODY></HTML>
Then, I can break the line between the HEAD tags, give myself an empty with Ctrl+Spacebar, use Ctrl+NP8, and type a title like “To Jeff, from Gene (November 14, 1998).” Then, I drop into the body, hit Tab to indent a line, type that date again, hit Return (or Enter) to get to the Margin, type Ctrl+P to get a paragraph line skip, type “Jeff,” as a salutation, and, in a new paragraph, indent and start the first paragraph. After each line (usually turn off word-wrap) I use Shift+Return to keep the indent for text. I use BLOCKQUOTE tags (in boilerplate or on the Tagset menu) to get the sort of indent I used for those lines above.
That’s the mechanical part. Most might assume that an e-letter is a web “page.” But that’s wrong. It’s wrong because most web pages are not written, but laid out. This is a very fundamental difference prior to the surface problem of most pages being Hallmark home pages, sales brochures, or examples of the “French whore house” school of design. The HTML formatting tags on the “engine” (manuscript) copy that drives the browser’s typesetting of the “interface” copies on screen or paper seem to people to be an opportunity to produce magazine pages. In an e-letter, which can be personal or business, but assumes an “as you think it” flow, or an e-memo, an e-report, an e-essay, or an e-article, the so-called formatting is punctuation, “blazing” the path you lay out for a reader, another who will follow your line, or system of lines, of thinking and imagining.
Why not use all the typesetting and layout skills developed for publishing to shape your “talk” to somebody, or anybody, about anything? Help chunk and sort what you have to say as you did in high school or earlier with commas and semicolons, but with all the devices available. Maybe a sentence or fragment in a paragraph isn’t positioned to be the topic sentence (or fragment) or the summary or the transition, but you want it slowed down as it passes through a reader’s “hearing.” So, you set it into italics even though it’s in a usual trough within the sequence of paragraph parts.
When writing poems in San Francisco in the sixties, I used white space punctuation counterpointed to the punctuation created with the usual marks. Dual readings happening not only simultaneously, but as parts of one reading, really. Shingled indents, line breaks in mid line or mid word, phrases “placed” in the page. It’s not too far removed to, when speaking data, to shotgun it into a table, into the cells, perhaps a table with no borders showing.
This is only a glimpse of an attitude. To see a result of writing rather than laying out a web scroll (because pages enter only when cut paper enters the picture), look at this FAQ scroll that you are reading. To see the mechanics of it, View Source ...in Notepad, if you do not have eWriter yet. (Your browser, asked for Source, may double space the actual lines of the source. That will damage the look. Allow for it.)
You can sense the writing in everything being aligned left. Even the example of a Table and its context which, on April’s library site is all centered. I’ve written that table and its accompanying buttons and label into a letter, here. It is a reference list on the site.
The general principle: An e-letter is written, not laid out. Since the writing may involve writing in tables and even things that come whole, like illustrations, the difference that the principle embodies is a subtle one, felt more often that it is easily described. To write, think, always, of tagging as punctuating.
Saving changes to bplates.set from Tools/Boilerplates.
Q. I’ve had trouble when I work on bplates.set using the copy of the file that Tools/Boilerplates brings up for me to use for importing boilerplate cylinders (seeing the set names). If I save, I get the SaveAs dialog and am told that my file name is New File. And sometimes if I close the file and then use Tools/Boilerplates again, I’m told that the file is already open.
A. The problem exists in 0.9m, and I never noticed it. In fact, I’m finishing up 0.9n and I’ve added a similar tool to the Tools menu for importing and exporting tagsets from the Tagset menu. While running it through its paces, I found this and then looked to see about the Boilerplates tool.
It’s fixed for 0.9n, which should be up before December 1st. My title bar caption of “Available Boilerplate Sets” where ordinarily an editor would have the filename is a big part of the problem. I didn’t change a reference. The work around is a bit cumbersome. In that SaveAs dialog, get into eWriter’s directory, double click on bplates.set or single click to get it in the edit and hit Save button. You'll be told file exists. OK the replacing of it. You will still get that “already open” message, so leave it open for the session. It will also be on the Recent Opens list, using one of the five slots, but if you don't save settings at the end of the session that will go away. It is fixed in 0.9n, due out by Decemner 1, 1998.
Upper one-byte characters and two-byte characters.
Q. I find that I can’t type the upper one-byte characters like è or © in eWriter. But I can in Notepad, and I can put them in text that I send and receive in an old Eudora-lite e-mail program.
A. I just checked that with Notepad and my own ancient Eudora-lite. The ANSI chars from 0128 to 0255, however, are not plain text. By definition plain text is 7-bit, not 8-bit, text. And I set up eWriter as a simple textwriter with no fancy export handling for a user to take care of. Since e-mail has to go out in 7-bit plain text, I suppose Eudora, possibly all major mailers, will run through the main text (not an attachment) and ...do what? Some coding. MIME? Well, what of BinHex receivers? Is there a standard? Could I count on that ...if I let my user type in those additional chars? And are there unsafe ones? I don’t find anything mentioned about a standard “encoding” in the main text.
I can see where this might become important for people using the European language keyboards and writing, say, in French or Norwegian. If the encoding was standard and as straight forward as it is in HTML, I could easily catch the key code coming in and substitute the coding. Or would that confuse the mailer programs? As you see, it’s an open question. I wouldn't mind some conversation on the subject.
As you know, for HTML use of the high one-byte chars, I have a lot of single character handling on the keys menu. Thinking about, say, entering those when the text is in a language using the accented characters, I could have a switch on the menu so that if the text is for HTML or XML and a 128...255 is issued by the keyboard, it goes into &#...; frame in the text. The switch is so that it does not go into regular plain text that way. I might end up with both. I’m thinking. I’d still have my menu’d items, of course.
I just coded a tentative solution for both of those keyings. I have two enablings on the Edit menu, only one of which can be enabled at a time and neither of which is saved across
sessions. One allows the high one-byte characters to be entered as text when entered from the NumPad and the other allows them to be entered as escape codes as is done using the Keys menu.
If you need this, say to write in French or Norwegian, email me of your need and I'll email back an 0.9n+ executable to use in place of your 0.9n until I have an 0.9o fully ready.
What about wide chars, the two-byte characters that are the foundation for unicode and allow 65K characters rather than 256? Delphi allows widechar handling. And Windows NT’s version of the Win32 API is friendly, but Windows 95’s version ...simply doesn’t work. Period. In any case, I don’t know what browsers do, or will do, if handed anything like a two-byte character. Presumably, in Windows 95 ...nothing. I guess unicode is something for the millennium. Two handlings for the upper one-byte chars, though, are in my thoughts. Conversation is welcome. In fact, it’s welcome on any answer in this FAQ as well as when used to bring in questions.
Why mailers that Render HTML do not render e-letters
Q. It occurred to me that your HTML letters don’t show up rendered in my mail client, even though Netscape’s Netcenter News messages--which are also HTML--do. I’m not sure what the differences are, but I’ve attached a recent Netcenter mailing. Take a look and see if there’s something in there you're not doing.
A. I (Fowler) send what I call e-letters. These are letters with HTML framing and punctuation set up to be very readable as manuscript in the text editor that most mailers use and then to be read in and printed from a (web) browser as typeset copy.
Now, some mail clients render HTML tagged texts in the typeset version so that the initial reading can be of the typeset version. Obviously, this is more convenient than doing a Select All and Copy, pasting in eWriter, saving as an .htm or .html file, and hitting Tools/Browsers for an automatic display. And presumably the mail client will, like a browser, have View source for “doubled” reading.
I'm pretty sure that adding this META line to the HEAD section will enable that rendering:
<META HTTP-EQUIV="Content-Type" CONTENT="text/html;
charset=iso-8859-1">
The skinny frame that has been
<HTML><HEAD></HEAD><BODY BGCOLOR="FFFFFF">
. . .
<BR></BODY></HTML>
that’s often expanded a bit to use the title to repeat a Subject line or an over-salutation like this:
<HTML><HEAD>
<TITLE>To Joe, from Gene (December 19, 1998)</TITLE>
</HEAD><BODY BGCOLOR="FFFFFF">
. . .
<BR></BODY></HTML>
Now, if you want a mail client (that can) to render the HTML'd version, add that HTTP-protocol MIME-type element to the header.
<HTML><HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html;
charset=iso-8859-1">
<TITLE>To Joe, from Gene (December 19, 1998)</TITLE>
</HEAD><BODY BGCOLOR="FFFFFF">
. . .
<BR></BODY></HTML>
For an e-letter which will usually be read first in its manuscript form (as most people do not upgrade to monstrous client programs), that skinny header is important. It goes along with indented paragraphs so that non-inline tags can be dropped into a gutter and other practices that aid the readability of the text. But, a recipient's convenience counts as well.
(I (Fowler) received the question and worked out the solution on a week-end, and my test mailing won’t be seen until Monday so, in effect, the solution is untested. However, that content statement is made for apps to use ...so I’m confident that these mail clients are using it. NOPE! Didn't do it. So I used ALL that the rendered missives used. A DOCTYPE, the HTTP-EQUIV line, and even a GENERATOR line that claimed use of MSHTML 4. Got back a response that even this was not rendered. I looked at all the examples. ONE THING seemed to be always there. A blank line just before the HTTP-EQUIV line. So, I added this in. But, my recipient is out of town for a week. I'll let you know. --gf, Jan. 11, 1999.)
Why isn't there a quick type for an XML Entity Reference?
Q. I’m doing some XML trials, and I find a missing menu item, though I have the 0.9o that just came with the new year (1999)?
A. Yup, I’m already making a list for 0.9p. For other readers, an Entity Reference is the way, in a document, to insert an entity declared in the DTD (Document Type Definition) file. The “entity” can be a string given in the declaration, or an entire file whose contents can be inserted. In HTML, and using the Keys menu, we use Character Entities, so that the “ampersand” will be inserted if the Entity Reference is “&” In XML, too, the Entity Reference will be the declared name with the ampersand before it and a semicolon after it.
The typing shortcut would be on the Tagset menu with those items above the area at the bottom where you will build a tagset. It would type “&|;” or the caret between the ampersand and semicolon. You’d type the name, use the right arrow to pass the semicolon, and go on typing.
The workaround... is to use Ctrl+G, and backspace. The Ctrl+G gets the “general symbol” from the Keys menu. This was for entering any ANSI number to get the Character Reference. To generalize it, backspace over (delete) the #. One extra keystroke ...temporarily.
Q.
A.
Gene Fowler
acorioso@ccnet.com
First posting: November 1, 1998
(continuous appending)