Text content (OCR)
_A2-Central’
A journal and exchange of Apple Il discoveries
formerly
Open-Apple
February 1989
Vol. 5, No. 1
ISSN 0885-4017
newstand price: $2.50
photocopy charge per page: $0.15
Fonts: on screen and in print
If you're going to be a citizen of the Apple {1 kingdom in the coming
decade, it's becoming clear that you're going to have to learn about
fonts. Once upon a time, only printers (Gutenberg's offspring, not the
ImageWriter's ancestors) had to know all about fonts, but those days
are now only a memory, like manual typewriter returns.
Although handwriting did once capture our written personalities,
this is 1989 and third-graders are desktop-publishing book reports.
Today you may feel that the typefaces built into your ImageWriter can
capture as much of your personality as you want captured, but tomor-
rows graphics-based personal printing will capture your imagination
as well as your personality. The future will offer you multiple opportu-
nities to express yourself—and to look like a clown if you don't know
all about fonts—so pay attention.
Fonts in printing, Traditionally, the word ‘font' has meant a com-
plete set of type (all the upper- and lower-case letters, numbers, and
punctuation marks) in one face, style, and size. This is exactly what
the word means in todays context as, well. However, instead of con-
sisting of wooden boxes filled with heavy lead castings, today's fonts
consist of files you can store on your disks filled with ethereal digital
images.
The 'face" or ‘typeface’ of a font refers to its basic design. In desk-
top publishing, this quality is also referred to as the font's ‘family’.
Some important elements of a family design include serifs (fine lines
that finish off the main strokes of a character, such as the little feet
and hats you can see if you take a close look at this M) or the lack of
them (called a ‘sans serif’ family) and the relative height of capitals,
lower-case letters (called the ‘x-height’), ascenders (the uprights on
‘d’, ‘h’, and friends}, and descenders (the legs dangling under ‘g’, ‘J,
‘p."q, and 'y’),
Traditionally, ‘style’ refers to whether a font is plain, bold, italic, or
bold-italic. Here in the age of desktop publishing, underline,
outline, shadow, and smaLt Cars are other styles you'll occasionally
see (very rarely, let's hope). In the days of lead castings, each style
(and each combination of styles, such as bold-italic) required a sepa-
rate font. On the Apple Il, styling is typically done by electronic
manipulation of a plain font.
Font ‘size’ fs traditionally expressed in a unit of measure called the
‘point’. There are 72 points to the inch. A font's size is usually consid-
ered to be the distance from the lowest descender to the highest
ascender. In print shops, the size doesn't include any of the white
space between lines. This white space is called ‘leading’ (rhymes with
‘red’, not "reed’), Thus, '10 on 12" describes a 10-point font on a |2-
point line~2 points of leading are inserted below each line of type.
Nowadays, the size of a font, like its style, can be manipulated elec-
tronically, however, the results are often unsalisfactory, particularly
when a small font is used to make a larger one.
Some typefaces or font families are designed primarily to be easy
to read. Other font families are designed primarily te attract attention.
When designing documents, please do the universe a favor and save
the attention-grabbing fonts for grabbing attention. If you set an entire
block of text in an attention-grabbing font, you lower both its readabil-
ity and the probability that anyone will read it.
Another easy document-design rule pertaining to typefaces is to
limit each of your documents to just a couple of font families. Most
printed material uses a serif typeface for text and a sans serif typeface
for headlines, or perhaps the reverse. Well-designed documents cre-
ate emphasis and flow by varying the size and style of the fonts, not
by changing font families.
While A2-Central isn't about to win any awards for graphic design
{too many large blocks of small type) we do stick to the same three
font families for everything we do—from the newsletter itself to pro-
motional materials to advertising to disk labels. The large "A2-Central*
at the top of our front page and the text you are reading now, as well
as our italicized answers to readers’ letters, are ail various sizes and
styles of a serif typeface called Benguial. Our headlines and the small
type used in figures and in the copyright/warranty box on the back
page are all various sizes and styles of a sans serif typeface called
Helvetica,
We also use a third typeface, Courier, when we want to show some-
thing as it would appear on your computer screen. Unlike most font
families used in printing, Courier is a non-proportional, or mono-
spaced, font. All its characters are exactly the same width. This means
that the letters in a block of text set in Courier will ine up in columns,
Just as they used to do on your typewriter and still do on your 80-col-
umn computer screen.
“While a single document should limit font usage to just a couple of
families, deciding exactly which families will create the look and feel
you want can he a big job. There are over 500 families available right
now for use with software such as AppleWorks GS and SuperFonts,
Font use on personal computers. The Apple 1] is the first person-
al computer alphabetically, chronologically, and in the use of fonts.
Early font usage on the Apple Il, however, was limited to screen dis-
plays. Later, software was developed that popularized dot-matrix print-
ing with fonts. Three early software packages that allowed the use of
fonts on the screen were Apple's DOS Toolkit, Higher Text II from Syn-
ergistic Software, and Apple Mechanic from Beagle Bros. Apple
Mechanic was the first program to offer proportional fonts (characters
of various widths) on the Apple Il.
Each of these three programs had its own special format for fonts.
5.2 A2-Central
This lack of standardization quickly became the standard. Publishers
of tater programs that allowed printing with fonts, such as Data Trans-
forms with Printrix and Fontrix and Springboard with Newsroom, each
designed its own unique font format. Broderbund, the exemplar
among publishers who make users’ lives miserable through copy pro-
tection and non-standardization, again led the pack by developing
two unique font formats, one for Print Shop and another for Print
Shop GS.
Fortunately, with the advent of the Apple IIgs, Apple created and
documented a standard font format. The format is based on the for-
mat of Macintosh fonts, only it's better. Not only all non-Broderbund
Iigs programs, such as AppleWorks GS, but also lle/IIc/ligs programs
such as Beagle Bros SuperFonts and TimeWork's Publish Ii!, use the
new format (or a slight variation).
A font clearinghouse. Because of the movement in the Apple II
community to a single font format, and because of the extreme con-
fusion in the Macintosh community about fonts, we recruited a sub-
scriber and GEnie user with a large collection of and interest in fonts,
Mark T. Collins from Waukesha, Wisc., to create an A2-Central Font
Clearinghouse on GEnie. What Collins is trying to do is bring togeth-
er a large library of public domain ‘Ilgs fonts that are bug-ree and
compatible with each other.
We've put a lot of Collins' work on this month's A2-Central disk.
including his shareware font editor, called Font Doctor. We've also
included a nice selection of public domain fonts, samples of fonts,
programs related to fonts, and two AppleWorks data base files that list
the specifications of the fonts currently in the GEnie clearinghouse
collection.
While I'm mighty proud of the work Collins has done for‘us on
GEnie, there are other sources of public domain IIgs fonts you should
be aware of. The other national online services, of course, include
many font files in their libraries, as do most user groups. TechAlliance
(formerly Apple Co-op, 290 SW 43rd St, Kenton, WA 98055) has sev-
eral public domain font disks available for $3.00 + .50 each. Beverly
Cadieux, 3103 Lake Stream Dr, Kingwood, Texas 77339 has put
together a set of six 3.5 disks with about 500 font files that she will
send you for $30.00 + $3.50.
In addition to public domain fonts, shareware and commercial
fonts have been developed for the Macintosh and some of these are
in the process of being converted to Apple II format. The only com-
mercial fonts originally developed under the [igs-standard that I'm
aware of were published by StyleWare. However, since Claris pur-
chased StyleWare that product has been in limbo.
TimeWorks also sells a font disk for Publish It! These fonts follow
the Ilgs standard except for one smaii detail—they have a file type of
$F7 rather than the standard $C8. To use standard IIgs fonts with
Publish it! you have to change the filetype to $F7. To use Publish It!
fonts with software that follows the standard, you have to change the
filetype to $C8. You can do this using disk zap software if you know
what you're doing. Or you can do it with a public domain program
sent to us by William Olsen of Riverside, Calif. The program is in our
library on GEnie in a file called CHANGE.FONT.FILETYPE.BQY and it's
on this month's A2-Central disk in a file called CHG.FNT-FILETYP.
At one time StyleWare also advertised that it would publish a font
editor for Iigs-style fonts so that each of us could create new fonts or
modify existing ones. However, that product didn't make it to market.
Beagle Bros is working on a font editor, but it isn't expected to be
ready until later in 1989, At this time, the only IIgs font editor we're
aware of is Collins’ Font Doctor.
Aspects of fonts. Neither computer monitors nor dot-matrix print-
ers are able to display fully-formed characters. What they do display is
very small dots, or ‘pixels’. The essence of a font file is a description
of which pixels are foreground (or ‘on’}) and which background (or
‘off) for each character in the font. If pixels, monitors, and sheets of
Paper were always perfectly square, life would be fair. But theyre not
and it Isn't.
While traditional printing-press printing uses what we consider fully-
formed characters, pixels aren't alien to the print shop. Photographs
are reproduced by ‘screening’ them, which converts each photograph
into a matrix of pixels, The term ‘resolution’ refers to the density of
the pixel matrix. The smaller each pixel is, the more of them a printer
can cram into a unit of space. And the more of them that get packed
in, the more ‘information’ there is and the better looking the graphic.
Vol. 4, No. 12
A typical printed photograph has between 120 and 150 multi-shade
pixels per inch both horizontally and-vertically. An ImageWriter, on the
other hand; has 72, 80, 96, 107, 120, 136, 144, or 160 black/white
dots per inch horizontally and either 72 or 144 dots per inch yertical-
ly.. Even though the ImageWriter, at 144 x 144, can match the resolu-
tion of a. printed photograph, its lack of multishadé dots means it
can't match the look of printed output.
While the Apple II was first with fonts, the Macintosh was the first
computer to use the Iigs-standard font format. It will be helpful
—believe me—if we look at the Macintosh a few minutes to under-
stand how we got where we are today.
For years the Macintosh was sold only with a built-in monitor and
all Macintosh monitors were exactly the same size (a little small).
When a Macintosh is properly adjusted, its active screen area is 18
centimeters wide by 12 centimeters high—a ratio of 1.5. Meanwhile,
the pixel resolution of the monitor is 512 wide by 342 high. Manipu-
late these numbers mathematically for awhile and you'll find that the
Macintosh screen has a resolution of 72 pixels-per-inch both vertically
and horizontally. Macintosh screen pixels are exactly one printer's
point wide by one’ printer's point high.
One of the design goals of the Macintosh was to creale a com-
puter that could display on its screen what it would print on its print-
er—what you See is what you gel, or WYSIWYG (wizzy-wig). The goal
was met not only in terms of fidelity to the proportions of the material
on the screen, but in terms of fidelity to its exact size. The Macintosh
screen displays a 72 x 72 pixel-per-inch matrix and the ImageWriter
duplicates that matrix exactly.
(Well, sort of. 1 own four ‘consistent-interface’ Macintosh programs.
Most of them have a standard item under the 'File’ menu called ‘Page
Setup’. MacPaint, which prints at either 72 x 72 (Print Draft’) or 144 x
144 (‘Print Final’) lacks the standard Page Setup choice. AppleLink
Industrial Edition prints using the ImageWriter's text characters and
although Page Setup appears in its File menu, it's never selectable.
Quark XPress, my desktop publishing program, and MacWrite both
have a selectable Page Setup. Among the items in the Page Setup dia-
log box for the ImageWriter (Page Setup dialog box contents vary
depending on what kind of printer you are using) is a goodie called
"Tall Adjusted’. When Tall Adjusted is selected (the default in Quark
XPress), you get 72 x 72 (Quality-Faster in the Print dialog box) or
144. x 144 (Quality=Best In the Print dialog box) square ImageWriter
pixels per inch. When Tall Adjusted is not selected (the default in
MacWrite}, you get 80 x 72 (Faster) or 160 x 144 (Best) slightly-tall
pixels per inch. If you import a perfect circle from MacPaint into a
document and print the document with Tall Adjusted off, you get an
ellipse. | suspect MacWrite has a Tall-Adjusted-off default so that it
naturally takes advantage of the ImageWriter's maximum resolution
(160 x 144) but I'm not at all sure this is reason enough to contami-
nate the Mac's otherwise perfect wizzy-wig design (notice that when
you flip the Page Setup to Tall Adjusted in MacWrite the ruler's inch
measurements are exact (if your active screen image is the specified
18 centimeters horizontally), while with Tall Adjusted off theyre not),
While we're on the subject, ‘Best’ quality printing on the Macintosh
gives you four times more resolution than ‘Faster’. When Best is
selected, the Macintosh automatically looks around for a font twice as
big as the one being printed. If it finds one, it uses it to increase the
resolution of the printed output. This trick increases sharpness with-
out changing the printed font size. Beagle Bros' SuperFonts does the
same thing and can match Macintosh output exactly on an Apple II (it
even has a Tall Adjusted toggle), but you don't get to see what your
printed page will look like while you're editing iL)
When we talk about printing fonts on paper or displaying fonts on
the’Macintosh monitor, we can talk about pixels-per-inch. When we
move to talking about Apple ff monitors, however, the perinch con-
cept drifts into the rasters, That's because Apple I] monitors come in
a wide variety of sizes, yet they all display the same number of pixels
horizontally and vertically. A "9-inch” monitor has smaller, denser pix-
els than a ‘19-inch’ monitor, but they both have exactly the same
number of pixels in total, The number of pixels per-inch depends on
the size of the monitor and becomes fairly meaningless. The variety
of monitor sizes also makes it very difficult to have any kind of fidelity
between the size of an image on the screen and the size of that image
when it Is printed. To accomplish that kind of fidelity you'd have to
January 1989
make sure that everyone had the same size screen (as Apple did with
the Macintosh).
Furthermore, the ratio between the height and width of the screen
image depends on the posilion of those small adjustment knobs hid-
den on the back or inside most monitors. The standard aspect-ratio
for Apple II monitors (and for television screens) is 4 units wide by 3
units high, a ratio of 1.333 {not the 3 x 2,.1.5 ratio of the Macintosh).
However, by turning the adjusting knobs, most monitors can be set
across a relatively wide aspect-ratio range. | suspect very few of you
have monitors that have been adjusted so that the aspect ratio of the
active screen area is exactly 1.353. This lack of standard adjustment
makes it virtually impossible to have any kind of fidelity between the
vertical-horizontal proportions of an image on the screen and the pro-
portions of the image when it’s printed. Every day, Apple I] users draw
perfect circles on their monitors and then print eggs. Now you know
why.
The problems created by the variety of monitor sizes and aspect
Tatios in the Apple II kingdom, however, pale compared to the prob-
lems created by the fact that the Apple II has more horizontal resolu-
tion than the Macintosh, but far less’ vertical resolution. The two
graphics modes usually used when llgs-standard fonts are displayed
on the Apple Il screen are double-high-resolution, which has 560 pix-
els across and 192 up and down, and 640-mode -super-high-resolu-
tion, which has 640 pixels across and 200 vertically. Notice that both
double-high-tes and 640-mode superhigh-res have more horizontal
resolution than the Macintosh (560 and 640 compared to 512). The
limitation of Apple Il graphics is the vertical resolution—just a little
more than half that of the Macintosh (192 and 200 compared to 342).
(The big limitation of Macintosh graphics, on the other hand, has
been the lack of color. A monitor's ability to do color, its vertical reso-
lution, and its cost are all related to one another. In the Apple il we
have a computer that can do color work on relatively inexpensive
monitors, but with limited vertical resolution.-The original Macintosh
could squeeze a lot of vertical resolution onto a relatively inexpensive
monitor, but it couldn't do color. Apple added color capability to its
Macintosh I] without giving up any vertical resolution, but for the price
of that machine's color monitor and video card you can get Apple's
lenied sesolution IIgs color monitor and have the Ilgs itself thrown in
for free.)
Ah, but what does all this screen-resolution mumbo-jumbo have to
do with fonts? The simple truth is that almost all of the IIgs fonts
available today were originally designed using the square pixels of the
Macintosh. By fiddling with the vertical- and horizontal-size knobs on
the back of your Apple II's monitor (Apple's own monitors don't have
horizontal-size knobs, but some others do), you can also get square
pixels on an Apple Il. However, square pixels are only possible in
standard-high-res {aim for an aspect ratio of 1.46). In the more com-
monly-used double-high-res and 640-super-hightes modes, pixels are
tail and skinny—usually a little more than twice as high as wide—no
matter how much you fiddle with the adjustment knobs.
When fonts designed for the square pixels of the Macintosh are dis-
played using the tall, skinny pixels on an Apple II, the once well-pro-
portioned characters become tall and skinny, too. (Characters that
originally looked tall and skinny on the Macintosh nearly disappear on
the IIgs.) With any given font you can actually get a few more charac-
ters across the screen on a Apple Il than on a Macintosh because of
the additional horizontal resolution. However, you can only get about
half as many lines on the screen, And those half-as-many lines get
stretched vertically so that they touch both the top and bottom of the
active screen area. That's why we have those tall, skinny characters
and pixels,
And it all boils down to this: you are sitting in a warm computer
room. You have a document on the screen in front of you. The docu-
ment uses a font originally designed for the Macintosh. You see tall,
skinny characters: You want to print this document, But do you want
your printed output to duplicate exactly what you see on your screen
(like the Macintosh) or do you want your printed output to have well-
proportioned characters (like the Macintosh)? What a dilemma.
If you own the standard Ilgs-issue AppleColor RGB Monitor and
want your printed output to look exactly like what you see on your
screen (a typical wish when using drawing programs, for example},
start by turning the knobs on ‘the back of your monitor until your
aspect ratio is about 1.44. According to the manual that comes with
A2-Central 5.3
that monitor, the active display area is adjusted at the factory to 20
centimeters wide by 15 centimeters high (the expected ratio of 1.33).
My own-.monitor's unadjustable width is only about 19.5 centimeters,
so I've set the height to 19.5/1.44 or about 13.5 centimeters. Perfec-
tion would be an active display area 20.3 centimeters wide by 14.1
centimeters tall. Mathematically this works out to 80 pixels per inch
horizontally and 36 pixels per inch vertically—exactly the resolution
the Ilgs uses when printing on the ImageWriter. Thus, if you can adjust
your monitor's active display area to these dimensions, what you get
on paper can exactly match what you see on the screen,
However, to turn "can" into "does" you have to understand three
options you get in the Ilgs Page Setup and Print dialog boxes. Like
righteous Macintosh software, desktop software on the llgs includes
Page Setup and Print choices in the File menu. The three critical
options are:
Orientation. Page Setup gives you an Orientation choice of Portrait
or Landscape. Portrait is what we'd normally call ‘normal’; Landscape
is what we'd normally call ‘sideways'. To get exactly 80 x 36 pixels
per inch on an ImageWriter you need to select Portrait. Because of the
limitations of the ImageWriter when you turn it on its side, Landscape
printing is done at 72 x 40 pixels per inch, which creates slightly
shorter, fatter characters.
Aspect ratio. Also hidden away in the Page Setup dialog box is an
essential item called ‘Vertical Sizing’. To create an exact duplicate of
what you see on your screen on your printer, you must set Vertical
Sizing to normal. But if you are printing clever documents using
fonts, you probably don't really want to print those tall, skinny charac-
ters you see on your screen, do you Bobo? To print text with fonts in
their ‘correct’ (MacWrite's 80 x 72) proportions, you must set Verti-
cal Sizing to condensed. Like the Macintosh's Tall Adjusted toggle.
Vertical Sizing flips between exactly what you see on your screen (nor-
mal) and better-looking text (condensed), There is no Tall Adjusted
toggle on the IIgs, just as there is no Vertical Sizing toggle on the Mac-
intosh.
Resolution. Finally, the figs Print dialog box lets you select between
‘Better Text' or ‘Better Color’. If you select ‘Better Text’, the IIgs will
look around for a doubie-sized font and print 160 x 72 pixels per inch
(normal) or 160 x 144 (condensed)—the equivalent of the Best Quali-
ty selection on the Macintosh. (If the selected font isn't exactly double
the height and width of the original, wave good-bye to wizzy-wig.)
The following table summarizes all this and a bit more:
ge printing resolutions in pixels per inch
Screen in 640-mode
portrait orientation landscape orientation
better color better text better color better text
normal, 80 x 36 160 x 72 Tax M4x 80
condensed 80 x 72 160 x Mid x 80 ‘Ma x 160
Screen in 320-mode
portrait orientation landscape orientation
better color better text better color better text
normal 40 x 36 x 72 36 x 40 ax 8
condensed 40 x 72 80 x 14d 6x 80 72x 160
Many subscribers have mentioned how slowly GS/OS programs
Print. It appears to me that the reason is that the GS/OS ImageWriter
printer driver always makes the ImageWriter print at its maximum hor-
izontal resolution of 160 dots per inch. If the size of the pixel to be
printed is larger than that, GS/OS will print several 160 dpi dots per
pixel. For example, you can see in the above lable that 640-made
portrait orientation in Better Color has a pixel resoiution of 40 x 36.
However, GS/OS actually has the ImageWriter print 8 dots (160 x 72)
of various cyan, magenta, and yellow combinations to create that
pixel. While “Better Text’ on the Iigs is equivalent to "Best Quality’ on
the Macintosh, ‘Better Color’ is not equivalent to the Mac's ‘Faster
Quality’. Once again, we find ourselves sacrificing something in com-
parison to the Macintosh (this time it's speed) in order to gain color.
The Iigs Print Manager uses multi-dot pixels with Better Color even
if you don't have a color ribbon. If you do have a color ribbon, be
sure to check the Color square in the lower-left corner of the Print dia-
log box. If you print in Better Color but don't have Color selected, the
3.4 Ad-Centrar
Print Manager tries to do gray-scaling with black dots rather than full-
color magenta-cyan-yellow mixes.
Font anatomy. Now that you understand all the various ways that
a font can be printed, lel’s look al an actual font file.
A llgs font file has four major parts, The first is the header, which
contains all kinds of interesting information about the font, The sec-
ond is the strike, which contains the pixel images of the characters in
the font. The third is the focation table, which has an entry for each
character in the font. The entry points to the character's pixel position
in the strike. The fourth is the offset/width fable, which tells how each
character is to be placed relative to the current ‘character origin’ or
‘pen position’ and how wide each character is. First we'll look at the
strike and at the tables, then we'll come back and investigate the
header.
The font strike. Two of the details we'll find in the header are a
pointer to the start of the font strike and the strike's row width. The
Strike is just a very long sequence of bits in bytes, but conceptually
the first bunch of bits in the strike represents the top row of pixels for
Figure 16-29
Part of a font strike
all the characters, the second bunch the second row, and so on. The
number of bits in a bunch for any particular font is the row. width. The
characters in the strike are all pushed together with no spaces
between them, so once you get the rows lined up, the strike looks
like Figure 16-29, which is unabashedly stolen from the QuickDraw II
chapter of Apple's Appie gs Toolbox Reference, Volume 2.
Each row of the strike is padded on the right end with enough extra
bits to make the row width end on a ‘word’ boundary (a ‘word’ is two
bytes}. Not every possible ASCII code has to have an image in the
strike, The header includes the ASCII code of the first (firstChar) and
last (astChar) characters in the strike; characters outside this range
are defined as ‘missing’. (Character codes can range from 0 through
255, however, the ASCII standard only defines the codes from 0 thor-
ough 127 and Apple has no recommendation other than to look at
existing fonts for 128-255.) Characters within the firstCharlastChar
range can also be "missing," as we'll see in a moment. Immediately
after the pixel image of fastChar in the font strike (at lastChar + 1) is
an extra Character that is used in the place of all missing characters.
Apple calls this character the "missing symbol" and has made it a
required element in the font strike (even if there arent any missing
characters). Usually it's either an empty box or an inverse question
mark, but sometimes a desiqner's copyright is crammed in here.
The location table. The location table is an array of bytes with a
two-byte (one-word) entry for each character code from firstChar to
JastChar + 2. The entry holds the distance, in pixels, from the begin-
ning of each row of the strike to the left edge of the character's
image.
To find the width of a character's image, subtract the location-table
entry for the character in question from the location-table entry for
the following character. The difference is called the ‘image width’. If
the character has no pixels at all, such as the space characler or a
control code, the image width can be zero (but it can't be negative).
For this scheme to work, the location table has to have an entry for
lastChar +2. This entry must point to the pixel just to the right of the
"missing symbol’, which is the same thing as holding the total length
of the font strike in pixels (ignoring any word-boundary padding).
The offset/width table. Like the location table, the offset/width
fable is an array of bytes with a two-byte entry for each character code
from firstChar to lastChar + 2. If the two bytes for a character hold
$FFFF, that character is missing from the font. Otherwise the first byte
of each entry holds the “character width," in pixels; the second byte
holds the offset. Both the width and offset can range from 0 to 254
pixels (not 255, because of the need for $FFFF).
‘Yol. 4, No. 12
The width is the distance, in pixels, that the ‘character origin’
should be moved (to the right only) after the character is drawn.
Some characters, such as SPACE, will have a positive character-width
(the character origin moves to the right when the character is typed)
but a zero image-width (no pixels in the character image). Other char-
acters, such as accent marks, may have a positive image-width (a
pixel image) but a zero character-width (the accent mark is drawn but
the origin isn't moved so that the next character typed will appear
under the accent). Other characters, such as control characters, will
have both image- and character-widths set to zero.
The offset tells QuickDraw how to position the character with
respect to the character origin. Characters can start either to the right
or to the left of the character origin. Since the offset is always a one-
byte positive number, however, we need an imaginary point some-
where to the Jeft of the character origin to start the offset from. This is
provided in the header by an entry called KernMax, which is the maxi-
mum distance, in pixels, from the left-most edge of the left-most char-
acter in the font to the character origin. If KernMax is negative, that
left-most edge lies to the left of the character origin; if zero, on the
character origin; if positive (unlikely), to the right of the character ori-
gin. To figure out where to position a character image, QuickDraw
adds the character's offset to KernMax. If the resulting number is neg-
ative, the left edge of the character lies that many pixels to the left of
the origin: if zero on the origin: if positive that many pixels to the right
of the orig
Figures 16-26 and 16-27 show a couple of sample characters from
an imaginary font. Notice that alt the lines and points that things are
measured from lie between pixels. The image width has to do with
the number of pixels in the character, 5 for the 'y', 7 for the “f; the
character width has to do with how far the character origin is moved,
7 for both the "y’ and the ‘f’. If, in this imaginary font, kermMax was -2,
the offset of the ‘y’ would be 2 and of the 'f', 1.
The header, The header of a IIgs font consists of three distinct
parts, The first part holds the name of the font. The second part is the
Jigs header. The third part is the Macintosh header.
The very first byte of a font fite tells how many characters are in.the
font name. The following bytes spell out that name. The Ilgs header
begins immedialely after that. The first two bytes of the Ilgs header
tell how far it is from there to the Macintosh header.
The Macintosh part of the header, as well as the strike, location
table, and offset/width table, are exactly like those in a Macintosh
font except for one detail: the high- and low-order bytes of two-byte
Charcterwicth
a
Character eA 3
eos
Trost as
foe 16:28 Chaco
Character with ro kerning L— Ascent ine
tLe! character
SO | so
en
sone in
Her crate ig
eg Le omcerine
aT
nage widths
Hgre 162
7
aractor kerning left
integers are reversed. This doesn't apply to the font strike, but the two
bytes in every other word in the Macintosh font need to be flip-
flopped. The origin. of this requirement isn't perversity, but the for-
ward-thinking architecture of the Apple II's microprocessor.
The first two bytes of the Macintosh section of the font are the Mac-
intosh ‘font type." which is ignored in the Apple II kingdom. The next
two bytes are the ASCII code of the first defined character, firstChar.
The next two are the ASCII code of JastChar, The next two are
widMax, the character width (in terms of origin movement) of the
January 1989
widest character in the font. The next two are kernMax. The next two
are the ‘negative of descent’, which ['ll tak about more in a moment,
If you took all the characters in a font and typed them one on top
of another with all their character origins in the same place, you'd
have a black mess. ‘The smallest rectangle completely enclosing this
mess is called the font rectangle’, according to Apple's toolbox manu-
al. The next two bytes give the width of the. font rectangle, in pixels.
The next two give the height of the font rectangle.
The next two bytes give the distance, in words, from this field in
the header to the beginning of the offset/width table. To get the dis-
tance from here to the beginning of the location table, you must sub-
tract 2 * (astChar - firstChar + 3) from the offset/width table distance.
The next two bytes hold the font's ascent, the number of pixels
from the font's base line to the top of the font rectangle. The next two
bytes hold the font's descent, the number of pixels from the font's
base line to the bottom of the font rectangle. And yes, ascent +
descent must always be equal to the height of the font rectangle. The
next two bytes hold the recommended leading (number of blank pixel
rows between the lowest row of one jine of characters and the highest
row of the following line) for this font. Applications may either use or
ignore this value. The ‘negative of descent’, disregarded earlier,
seems to be disregardable. There's little relationship from font to font
in the number stored here. Some have -descent, some -ascent, and
some some other number.
The next two bytes give the width of each row in the font strike, in
words. The next byte is the start of the strike. The length of the strike
is the row-width multiplied by the height of the font rectangle. The
location table follows the strike, and the offsel/width table is at the
end of the font file.
The Ilgs part of the header adds several pieces of information
that were left out of the Macintosh header. As mentioned earlier, after
the embedded font name comes a two-byte offset, in words, from this *
field to the Macintosh part of the header. In the current version of the
igs font format (version 1.1), this field holds a 6. Apple may expand
the Ilgs font header in the future to add additional information. Soft-
ware that uses this field to find the Macintosh part of the header, how-
veh will be totally compatible with any future versions of the Igs font
format.
The next two bytes of the ligs header hold the font's family num-
ber, All fonts in the same family (i.e., all Geneva fonts or all Helvetica
fonts) shoutd have the same number here. On the Macintosh, one
byte of the ‘font type’ is used for the family number and the other is
used for font style (plain, bold, italic, etc.}. This means only. 256 fami-
ly IDs are available on the Macintosh, but there have been far more
font families than that for years already. Overlapping font IDs are a
real problem in the Macintosh world. Here in the Apple Il kingdom, on
the other hand, we have 65,536 available IDs. It's important that all
the fonts you use have matching IDs if they are in the same family
and unique IDs if they are in different families. If they don't, the print
manager can get mixed up and use the wrong font.
One of the major reasons we started the A2-Central Font Clear-
inghouse on GEnie was to keep our kingdom's font {Ds organized.
Apple Developer Technical Support doesn't assign [Ds to individual
fonts. Instead, it assigns ranges of numbers to companies or groups
who distribute fonts. The groups, in turn, can assign numbers within
their range to individual font families as they choose. Table | is a list
of ID assignments as of mid-January 1989. The A2-Central Font
Clearinghouse, for example, has been assigned 2,048 font-family ID
numbers between $7C00 and $83FF.
While most of the assigned ranges have gone to commercial com-
panies, we intend to reassign the numbers in our range to fonts
designed by individuals. You can get your own font ID number from
us simply by designing a new font family and allowing us to put it in
our library. We accept either public domain or shareware fonts.
One interesting bit of information about a font family is whether it
is designed to be displayed at the Macintosh square pixel 1:1 ratio or
at the Ilgs tall-skinny pixel 11:5 ratio, While condensed printing solves
the tall-skinny character problem when you're printing 1;1 fonts, pro-
grams that are screen-oriented rather than paper-oriented need some
fonts that are designed to look fat and happy on the screen. We'ré try-
ing to give new 1:1 fonts family numbers below $8000 (32767) and
new 11:5 fonts family numbers of $8000 and above. Of existing
fonts, most of Publish it’s fonts and a few of Styleware's are 11:5.
A2-Central
Table 1: Apple Tgs Font Fanily Number Assignonts, January 1989
DECIMAL iad
~ 1:1 Ratio (Macintosh format fonts) -
Q- 127 00- 7F Apple's Macintosh fonts
126- 255 80- EF vendor Macintosh fonts
240- 255 FO- FF also used by Stylevare (Claris)
256- 991 100- 3DF ~reserved for future assignment-
992- 1055 380-420 -asigned to Styleware (Claris)
1056- 8191 421-1FFF -reserved for future assignment—
8192- 8447 2000-20RF -assigned to unknown vendor
8448-27903 2100-ECRF -reserved for future assignment
27904-28159 6D00-GDFF -assigned to unknown vendor
28160-31743 6200-7BFF reserved for future assignment—
31744-32255 7C00-7DFF GEnie A2-Central-New I1gs Families-1:1 Ratio
32256-32767 7E00-7FFF GEnie A2-Central-Macintosh Conversions
~ 11:5 Ratio (640 SR format fonts) -
8000-83FF GEnie A2-Central-New IIgs Families-11:5 Ratio
33792-65279 8400-FEFF -reserved for future assignment
65280-65535 FFOO-FFFF Apple's IIgs Fonts
The third two-byte word in the IIgs section of the font header
denotes the font’ style. The algorithms used to electronically change
plain fonts into bold, italic, and so on don't work well with some
fonts. The style byte allows the IIgs to have prestyled fonts, Fonts that
are prestyled are not styled further—for example, if you have a
prestyled italic font and italic checked in your Style menu, the italic
font isnt italicized a second time electronically. It could still have
bold or other styles added, however. The meaning of the bits is shown
in the following table:
Meaning of bite in the font style byte:
5.5
32768-33791.
7654 #3210
0 0 0 | | | | Bold (Plainsall bits 6)
reserved | | | Italic
|| Underline
| Outline
Shadew
Only the low-order byte of the style word has any meaning. A*1' bit
means the style has been applied. A byte value of zero means plain
style. The bits are cumulative, for example, $0011 means bold-shad-
ow, On the ligs, underline requires a descent of a least three pixels.
That's why the Ilgs system font, Shaston 8, won't underline.
The fourth two-byte word in the IIgs section of the header gives the
point size, On the: Macintosh, point size is given as part of the file
name. Since Macintosh screen pixels are, in fact, exactly one printer's
point high, one would expect Macintosh fonts to be given ‘sizes*
exactly equal to their font height in pixels. But when you examine the
headers of a number of fonts, you find that this is the exception
rather than the rule. In some cases, it appears that the font size was
set equal to the ascent, which, while non-standard, does have at least
some meaning.
In the majority of fonts, however, there is little relationship
between ‘point size’ and the actual pixel height of the font. For exam-
ple, our library's collection of fonts claiming to be 9 point actually
have pixel heights ranging from 5 to 14, with stops at all pixel heights
in between. Fonts claiming to be 24 point have actual pixel heights
ranging from 13 to 40 and beyond. While there is some point-size
meaning within families (a 24 point Shmoe is almost always about
twice as big as a 12 point Shmoe), between families there is almost
none.
Even within families, fonts that are billed as exactly twice the size
of smaller fonts are often a few pixels off. If you select a high-quality
printing mode that looks around for a double-point-size font, your
entire document can be reformatted when the double-size font isn't
really double size. Because of the mish-mash of point sizes we've
inherited from the Macintosh, what you see may not be what you get
at all. The: lack of relationship between stated point sizes and the
actual pixei height of fonts is detrimental to the mental health of our
community.
The fifth two-byte word in the llgs section of the header gives the
version number of the font format. All fonts that have the 6-word IIgs
5.6 A2-Central
header discussed here should have $0101 (v1.1) stored in this word.
The sixth and last two-byte word in the ligs section of the header
gives the ‘font bounds rectangle extent’. This is the farthest possible
horizontal distance (right or left) from the character origin to any pixel
that can be altered by drawing in any character in the font. If this
value is smaller than it should be, QuickDraw II can overwrite memo-
ty locations that don't belong to it and create all kinds of havoc. This
value must always be equal to or greater than both kenMax and
widMax (the largest character width in terms of origin movement). If
any character in a font kerns to the right of the most distant new
character origin, this value will be larger than widMax.
In addition to the problems with point size and extent we've just
discussed, there are a number of other bugs lurking in font files. For
example, some fonts have entirely blank rows of pixels at the top or
bottom of the font. This ‘forced leading” is against the rules. Others
have impossibly large kernMax and widMax values.
One of the primary tools I've used to learn all this is the Apple-
Works dala base file Mark Collins constructed that contains all the
header information from the fonts in our clearinghouse. First he
wrote a prograin that dug the values out of the headers and put them
ina text file, then he converted that file into an AppleWorks data base
file. This file is available on this month's disk and in our library on
GEnie. You can use it to find out, for example, the true pixel height of
a font. You can use it to find out how many font sizes are available in
a family and whether the larger fonts are really exactly double the
size of smaller fonts.
That's all | Know about fonts so far. | expect to learn more as the
months go by and our font clearinghouse grows. Meanwhile, if you're
looking for exactly the right font to catch your personality on paper,
Vol. 4, No. 12
Miscellanea
Computer Systems News, ‘The Newsweekly for the Computer and
Systems Integration Business’, lead off its year-end issue with a list of
things that didn't change at all in 1988. At the top of the list was:
THE APPLE Il: No matter how many new Apple Ills, Lisas, and
Macintoshes the company introduces, no matter how many times
experts predict its demise, Apple's classic Apple 1! continues to
flourish. It's worse than crabgrass!
From the Wall Street Journal's "Business Bulletin", December
15, 1988:
COMPUTER SALES stay strong despite a predicted slowdown,
retailers say. The big buyers: small and medium businesses, as
well as people who want computers for the home...
And from Apple's 1988 Annual Report:
Our Apple I! product family continues to be the most popular
choice in the primary and middle schools, with an installed base
over two million strong....Though we think of Apple [i computers
primarily in relation to schools, we found in 1988 that many fami-
lies and small businesses are also buying the Apple ligs, because
of its all-around strengths as a personal computer. Not only can the
Jigs run more educational software than any other personal com-
puter, it can also run general productivity software—everything
from word processing and home finance to small-business
accounting.
Apple's December Dealer Service Notice reminded dealers that
you know where to look.
Ask
Subscriber Tom Weilshaar of Overland Park,
Kansas called in a correction to my statement
in the January issue, last paragraph on page
4.94, that AppleWorks 2.1 will ‘use as much
memory as you have" on standard-siot memory
cards, In fact, Weishaar said at the end of a
Jong day of troubleshooting that followed a sys-
tem reconfiguration, AppleWorks 2.1 immedi-
ately crashes if you have a standard-slot card
with more than two-megabytes on it—even if
there's less than 2 megabytes of free space. As
soon as AppleWorks sees all that memory it
hangs with its tongue drooping out.
GS/OS parallel driver
Td given up hope of ever being able to print
anything from a SYS16 program on my parallel
Epson FX-85 connected to a SLOTBUSTER Il
card in my ligs. Even the new parallel driver
that comes with GS/OS bit the dust.
It turns out that the driver does an ID byte
check of the parallel card. Randy Carlstrum at
RC Systems, the company that makes the Slot-
Buster, solved the problem. Bload PARALLEL.
CARD,A$2000,T$BB and check $22FD. The
original value is $14. Change it to $31 for the
StotBuster II. Since the SlotBuster emulates the
Super Serial card, this may work for other Super
Serial card look-aikes too.
Tuckerman Moss
Orinda, Calif.
When I got my copy of GS/OS I was very
excited about the parallel printer driver. |
thought my two-year wait was over and | could
finally use my Apple Dot Matrix Printer and
Apple Parallel Card with these. great new ligs
programs. However, all | got was a message say-
ing that the driver didnt recognize Apples own
parallel card. I paid over $1,300 Canadian in
1983 for this printer and card and believe there
must be an alternative other than selling it off
for $200 and buying an ImageWriter.
Lee Luker
Brandon, Man.
! don't know for sure, but |. suspect Apple's
driver can be made to work with any parallel
card that follows the Advanced Firmware Proto-
col (also known as ‘Pascal 1.1"; see-'Controk
Interface) S(tandards)," October 1987, page
3.65 and ‘Using the Advanced Interface,’ Jan-
uary 1988, page 3,92) by patching the ID byte
routine.
Apple's own parallel card doesnt follow this
protocol, however. if | was positive that the
Apple Dot Matrix printer could emulate all the
necessary Image Writer functions ‘(I think it can)
I'd recommend changing paralle! cards and
keeping your DMP. Remember, however, that it
prints more -slowly than an ImageWriter and
people are already aghast at how long it takes
to print on the imageWriter using the GS/OS
drivers.
Some stuff about Siders
1 just had to replace the power supply in my
10 meg Sider I. I found the PS-ASTEC from JDR
Micro Devices was a perfect fit electrically and
physically and it only cost $25 as compared to
$150 from First Class Peripherals.
Craig Peterson
Santa Monica, Calif.
[have an older 10 meg Sider that first boots
DOS 3.3 and then loads and runs ProDOS 16,
Now | want to use GS5/OS. What options do |
have other than buying a new hard drive?
Rich Katz
Los Angeles, Calif.
You can’ get an.allProDOS ROM for your
Sider Controller card for $59.95 plus shipping
from Advanced Technical Services, P.O. Box
920415, Norcross, GA 30092 404-441-3522
(GEnie email address is STEVE.PARK). You'll
have to reformat the Sider after installing the
ROM, You'll get one large ProDOS volume that
GS/OS is comfortable with.
Some stuff about SCSI
The Apple SCSI card seems to be incompati-
ble with hard drives that use the Adaptec SCS!
controller. The card has no problem talking to
the drive. | can read and write blocks to any
partition. But if you ask the card for the status
of the partitions, even numbered partitions are
ok while odd numbered partitions return status
byte $00. I also have a SCSI card made by
C.M.S., It has no problem talking to the drive,
but my drive is larger than 64 meg, which is the
limit of the C.M.S. card, so | am interested in
using the Apple card to get the full capacity of
the drive.
Jock Cooper
Gallatin, Tenn.
After reading ‘An introduction to SCSI’ in
your December issue (pagé 4.85), | was still
confused as to whether the Apple SCSI car¢
would work properly with my drive. | have a
C.M.S. SD20/A2S (stackable). So | decided to
get one and try it. I'm pleased. to say that it
works flawlessly. The GS/OS thermometer
works correctly. The Get Info button in
January 1989
part 341-0437, Rev A is the current ROM chip for the Apple {I SCSI
card, that it's required for GS/OS, and that customers should get free
upgrades. Apple's Profile interface card also needs a new ROM to run
GS/OS. This one is part 341-0299, Rev B, and is also part of a free
upgrade program. Appie has also provided dealers with a new system
diagnostic disk for the Apple igs (v3.1) that can identify RAM chips
that aren't ‘CAS before RAS,’ as required by that computer. Apple's
previous diagnostic programs didn't test for this and failed to identify
this not infrequent cause of ghostly problems.
Timeworks has released Publish It 2, a more advanced version
of its original desktop publishing program for the Apple lle, lic, and
Igs. The advanced version allows larger documents by including sup-
port for extended memory cards, can print on all PostScript printers,
can use Print Shopcompatible graphics, and has several additional
features not found in the original program. The original version is still
being sold for $99.95, however. The new version is $129.95.
Upgrades for registered users of the original Publish it! are available
for $30 from Timeworks, 444 Lake Cook Road, Deerfield, IL 60015,
312-948-9200,
GS/OS tips: you can't boot GS/OS by executing the PRODOS file
on a GS/OS system disk. It has to be a power-on or three-finger salute
reboot. If the Finder's constant clicking on 5.25 and UniDisk 3.5
drives bothers you {its looking to see if you've switched disks), open
up your SYSTEM folder, then your DRIVER folder, and click on the driy-
er for the disk that's bothering you. Then press operi-apple (Get
Info). In the lower-ight corner of the information dialog box is an
‘inactivate’ selection. An ‘inactivate box" appears for anything in your
DRIVERS, DESK.ACCS, or SYSTEM.SETUP subdirectories, but you have
A2-Central 5.7
In the envelope with this month's newsletter you'll find our
annual Index rather than a product catalog, so I'm going to squeeze in
what's new at A2-Central this month right here.
Two new Apple Ilrelated books have just been published. Brady
Books has Programming the Apple IIgs in Assembly Language by
David Eyes and Ron Lichty. Eyes and Lichty also wrote what many
consider to be the Bible of the microprocessor in the ligs, Program-
ming the 65816. Their new book is $29.95; if you order from us our
item code is $S-005.
Addison-Wesley is about to release Using AppleWorks GS, by Doug
Brown. Brown was a member of the AppleWorks GS documentation
team. His book's discussion of each module of AppleWorks GS
includes a description of basic features, shortcuts, and sample appli-
cations. This book is $19.95; if you order from us our item code is
AW-027.
Cirtech has raised the prices of its memory cards that use 256K
RAM chips, which continue to be difficult to find. We'll sell our current
stock of cards and stick by the prices in last month's catalog till
February 28, but expect to see some price increases after that.
We're taking over Kansas City’s Avila College campus on the
weekend of July 21-22-23 for an A2-Central Developer Confer-
ence. For one slightly extravagant fee you'll get a conference teeming
with Apple Il hardware and software developers, all your meals,
and—if you're at the head of the line—free dormitory accommoda-
tions. Apple has already committed to send us some engineers, evan-
gelists, and support types and to throw a party, so plan now on com-
ing to KC this summer. More details next month.
to reboot GS/OS to get your new selection to take effect.
Advanced Disk Utilities returns 'SCSI.1* instead
of ‘DEV.1* as before. And the partition button is
no longer dimmed.
Ron Kunkel
Machesney Park, III
I'm writing to add to December's discussion
of generic hard drives for the Apple il. I built my
own SCSI drive and have pertinent information
on my BBS.
| ordered a Seagate ST277N 65 meg drive
mechanism from Hard Drives International
(1208 E Broadway Rd, #110, Tempe, AZ 85282,
602-784-1038, 800-234-DISK) for $449, a beau-
tiful case with 30 watt power supply, fan, and
all cables from Tulin Corp (2393 Qume Drive,
San Jose, CA 95131 408-452-9025) for $119
and a C.M.S. SCSI card from CDA Computer (1
CDA Plaza, P.O. Box 533, Califon, NJ 07850,
201-832-9004, 800-526-5315) for $115. Hard
Drives International also offers smailer capacity
SCSI Seagate drives for $319 (20 meq) and
$419 (40 meg).
The Seagate ST277N consumes approximate-
ly 13 watts of power. The Tulin case can hold
two drives and its 50 watt power supply is more
than sufficient to power two drives. Although
cheaper cases can be had, then usually don't
come with all the necessary hardware, such as
cables, which may be difficult to find otherwise.
The Seagate ST277N mechanism is what
C.M.S. used in my C.M.S. SD60 A2S drive. I
selected their card rather than Apple's because I
was concerned with compatibility with my exist-
ing CMS. drive.
Anyone wishing to enter the world of SCSI
hard drives would do well to build their own at
considerable savings. Construction is quite sim-
ple (if | can do it, anyone can), The drive mech-
anism attaches to the bottom plate of the case
with four screws. One of the two plugs from the
power supply connects to the back of the drive
mechanism, as does the cable from the DB-25
connectors on the back of the case. A cable
coming from the drive plugs into the power-on
Tight on the case. And the top of the case is
reattached using three screws. That's all there is
to building it. Total construction time was about
10 minutes for the first one | built, five minutes
for the second...and that included a two minute
coffee break.
If anyone needs more help, they can contact
me on my bulletin board, The Washington
Towne Crier (201-689-3649, 300/1200/2400)
and I'd be glad to assist.
Dr. Kenneth Buchholz
Washington, NJ.
Back in December, page 4.86, last para-
graph in the first column, | said that C.M.S.
SCSI cards don't work with certain other cards
in the computer. C.M.S. has discovered that
the source of the problem was a batch of
defective Mitsubishi chips. The bad chip is the
biggest chip on the card—C.M.S. will give you a
free upgrade if you have a chip from the defec-
tive batch.
If you add a generic SCSI drive to your'sys-
tem using an Apple SCSI card and you run
GS/OS, remember to use the Installer to add
the SCSI driver to your System Disk. Until you
do that, GS/OS won't recognize the drive.
Control Panel straightener
My wife teaches biology at a schoo! that uses
many Apple Ils. One problem they are having
with the ligs is that they waste a lot of time trou-
ble’ shooting and correcting altered control
panel settings. Is there a way to lock the control
panel so that students cant change the set-
tings?
Allan Seidel
Olivetter, Mo.
You could play games with configuration
programs that might slow the students down,
but any student worth his or her own computer
will quickly figure out how to work around it.
The easiest solution is to boot the !igs with
option/control/reset (this is the standard
three-finger salute, but hold down the option
key instead of open-apple.) This brings up a
menu with four choices. Choose the second
‘one, ‘Set system standards and 60 hertz." This
will reset the entire IIgs contro! panel to nor-
mal (checkmarked) settings.
Rock around the clock
If you use Apple's System Disk 3.2, keep the
ALARM.CLOCK NDA out of your SYSTEM/
DESK.ACCS folder. With this NDA installed, your
SYS16 software will hang when you try to print.
If you use GS/OS, keep the MENU.CLOCK file
out of your SYSTEM/SYSTEM.SETUP folder. This
file, which inserts a clock into your menu bar,
won't let you run your ProDOS 8 software but
will drop you into the monitor or hang your
computer each time ProDOS 8 starts.
Luigi Bruno
Rome, Italy
Cassette in, sound out
This machine language program will allow II-
Plus and [le owners to use their Apples as tape
players (for reasons unknown to me). Just con-
nect the recorder to the cassette connector on
the computer, put in a nice cassette, press play,
and run the program. Adjust the volume for
best results. You will hear the tape through your
Apple speaker. It works best with simple
melodies, Here's the program:
6000: AD 60.CO LDA $C060
6003: 30 FB BAT $6000 (or EPL—no difference)
6005: AD 30 CO LDA $0030 {any even number of
6008: AD 30 CO LDA $c030 WDA §C030s)
600: 4c 00 60 gMP $6000
Ron Maimon
Syracuse, N.Y.
What your program does is monitor the
‘state’ of the cassette input ($CO60—it can be
‘on’ or ‘off’). In one of those states it togglés
5.8 A2-Central
the speaker output ($C030). Here's another
program you should try:
6000; AO I DY $1E Create 180 microsecond
6002: 88 bey delay by counting
6003: DO FD BRE $6002 down
6005: aD 60 CO LDA §COE0 Get cassette input
6008: 45 2F © BOR §2F compare it to LAST
600A: 10 F4 “HPL $6000 if no change, restart
600C: 45 2F BOR $2F Input changed. Fix A
600: 85 2F © = STA S2F store for LAST compare
6010: AD 30.CO LDA $C030 Tickle speaker
6013: 4C 00 60 UMP $6000 Start over
This code segment comes from an article
called ‘Your Apple Can Talk’, by Bob Sander-
Cederlof, which was published in the Novem-
ber 1982 Apple Line. This code
segment only toggles the speaker output when
the state of the cassette input changes. In the-
ory, it should provide a cleaner sound than
your program, but I havent actually tried either
one.
You can use a microphone as well as a cas-
sette player as your input device. Using the rest
of the program from Apple Assembly Line,
you can ‘record’ and ‘playback’ about 10 sec-
onds worth of sound samples. The quality is
scratchy and low-fidelity compared to what can
be done with sound on the ligs, but it’s an
interesting place to begin a study of computer-
generated sounds.
Option-key watchdog
In your answer to January's ‘Foreign Accents,
cont’ you mention the possibility of a desk
A2-Central’
z
© Copyright 1989 by
A2-Central
Most gis reserved. All prograns pubished in A2-Centrat ae oubic
comain ard may be ceped and dried witout charge. Aol user tous.
and sonar ors may oan permission repr ars fom te are
by spocte witen request
\Wirten, ected, and published by
Tom Weishaar
with help from:
Dennis Doms Sally Dwyer Dianne Plumberg
Tom Vanderpool Steve Kelly Joyce Hammond
Dean Esmay
A2Ceatral~ited OpensApple to.gh January. 1983—4as been pub
lished monty sco vanvary 1965. Werd-wal prices (ir US. doers, ame
debvery nuded at ro anion char) $28 fo year: SS or 2 years: $78 tor
‘9 yeas, Al ban sues ar cute aalable fx $2 each: bound, indexed ec
tons of cur ist tree volurmas are $1495 each. Voumes end wi fe Jaruary
lesan dex ‘ore prior velume iced weh Ie February ssve,
The ul et of each issue of A2-Centratis avalatle on 35 d sks, tong
with & selection ofthe best new pubc domain and shareware fles and pro-
grams, or $84 a year [newsletter and dsk combined), Single disks aro $10.
Pease sod al crrespordence
A2-Central
P.O. Box 11250
‘Overland Park, Kansas 66207 U.S.A.
‘A2-Central isso in an unprotected fama! for your corwaniance. You
ae encouraged Io Take back-up archival copies of easy-to-read eniaged
Opes for your on use witout Charge. You may also copy A2-Cemtra ot
<dstribwion io aes. The distbuto fe Is 15 cers per page per copy ds
ribuied.
WARRANTY AND LIMITATION OF LIABILITY, | warrant hal most of tie
Itormaton in A2-Cemtral is usoht and comet, athough drivel and mi-
takes afe included fom time lo Ume, usualy unitentonaly. Unsatisieg
subscribers may carcel tilt eubscrgton al any time and reco‘ a ful
telund of their fast subsergton paymert. The unflled portion of any paic
subscripion wil be retundod even to satisfied subscribers upon request
MY LIABILITY FOR ERRORS AND OMISSIONS IS LIMITED TO THS
PUBLICATIONS PURCHASE PRICE. ino case shall | or my conlrbulors
be lable lor any incident) or consequential damages, not for ANY dam
agesin exons 0) the fs paid by # subsorber
{SSN 0865-4017
Printed inthe U.S.A
GEnie malt: A2-CENTAAL
919-469-6502
accessory being an ‘option-key watchdog’ for
programs that don't do it themselves. Our
MECC Key Caps disk accessory, which you men-
tion earlier in the same paragraph, will also
function this way. It allows using option, option-
shift, and control-key combinations to access
any character in a font from within an applica-
tion. As far as we know, it works with all appli-
cations, including AppleWorks GS, that doesn't
use the option key for other purposes. For
those that are already using the option key, you
can use the MECC Key Caps desk accessory to
type the character and then cut or copy it to the
clipboard and paste it into the application.
As you noted, MECC Key Caps is included
with the purchase of our Calendar Crafter pro-
gram and will be included with our other Iigs
products in the future.
Paul R. Wenker
MECC Technical Services
St Paul, Minn.
65802 & No-Slot Clock
I recently put a 65802 in my lle and found
only one incompatibility problem. The software
that came with the No-Slot Clock would no
longer recognize the clock or even confirm that
it exists,
Has anyone else confronted the problem? Is
there a solution?
Frank Eddy
Ogden, Utah
Someone out there knows.
RAM chip choices
I have an Apple IIgs with an Apple memory
expansion card and am considering expanding
it to accomodate AppleWorks GS and other pro-
grams that require at least 1.25 megs of RAM.
In making a choice of which card to buy, one
of the most difficult choices is whether to pur-
chase a card that uses 256K or 1 meg chips. I'd
appreciate your thoughts on which is the better
choice based both on price and continued avail-
ability. I'd also appreciate your thoughts on
when the best time might be to buy additional
RAM given current trends.
Scott Lorigan
Elk Grove, Calif.
My opinion is that you should be buying
cards that use I-meg chips. On a per-mega-
byte basis, these chips are already cheaper
than 256K chips, and I expect the gap to widen
from now on. The 256K chips are getting hard-
er and harder to find. The |-meg cards are far
more expandable, are cooler, use less power
and should be more reliable.
The ‘best time’ to buy is when you need
them. | expect the price of I-meg chips to start
going down within a few months, but the drop
will be gradual . They'll always be cheaper
tomorrow, but if you always wait till then to
Vol. 4, No. 12
buy, you'll never get any.
Pascal and 3.5 disks
1 am unable to transfer/copy Pascal files from
5.25 disks to 3.5s. After copying, it gives me
the message UNBOOTABLE DISK when the copy
is booted.
R, Naidu
Te Kauwhata, New Zealand
Assuming you are using Apple Pascal 1.3
(earlier versions don't support 3.5 disks), first
format your 3.5 with the Pascal 1.3 Formatter
utility. If you format it with Apple's System Utili-
ties or most other formatting programs, the
disk wont boot. After formatting, copy your
Pascal files to the 3.5 using the Pascal Filer.
Source for Curriculum Guides
Please send more information on Apple's
Curriculum Software Guides, which you men-
tioned in Decembers Open-Apple (page 4.85).
I am particularly interested in the Foreign Lan-
guage K-12 volume.
Richard Melpignano
Bellingham, Mass.
Those books are published by Apple itself
and are consequently available only from
Apple dealers who care enough to stock them.
If your dealer doesnt, call directory assistance
in a few large cities near you until you find
Apple's regional sales office. Then ask them for
the name of a dealer that stocks the books.
Lam is back
Here's an update to ‘igs missing CALL -144"
in your August 1987 issue, page 3.53—CALL -
144 is alive and well in the ROM version 01 of
the Ilgs. In fact, it is officially documented by
Apple as ‘MONZ4" at the top of page 256 of
Apple HGS Firmware Reference.
However, the Lam routine still doesn't work if
you use $D9C6 as the ‘return to Applesoft’
address. You have recommended this address
in the past (‘Picking Up Applesoft,’ February
1985, page 1.12; ‘A charming difference,”
March 1985, page 1.23) because on classic
Apples it work from within Applesoft subrou-
tines.
On the Iigs, you have to use $D7D2 {or
$D823, which jumps to $D7D2) as the return
address. This means Lam is back, but not from
within subroutines, The Lam routine will also
crash if you run the BASIC program after setting
a non-zero bank in the monitor. Prefixing the
$0300 with a bank byte fixes this problem.
Here, for example, is Lam alla Kashmarek
(March 1986, page 2.16) modified so it will run
on the Iigs:.
10 ; REM Lam a'la Kashmarek/Robbins (IIgs)
100 BELLS = CARS (7)
$0 c§ = "90/0300:20 7B DD 20 00 B6 AA AD 00
BL SE 09 6099.00 O2c8ca
DOTS A2 04 8648 4¢ 70 F
¥ 03F6:00 03 w D7D26"
200 PRINT BELLS;*Start'": REM Standard Lam
210 FORT= 10 LEN (CS)
PORE 51141, ASC(MIDS (C8,1})4128 :
HEMT : PORE 72,4: CALL -144
220 PRINT BELLS : REM Lam ala Kashmarek
230 & Cg
240 PRINT BELLE
250 ED
Bill Robbins
Osaka, Japan