pkg://vogl.tar.gz:336155/
vogl/
docs/vogl.3
downloads
.TH VOGL 3 "12 Oct 1993" "VOGL 1.2.8"
.UC 4
.SH NAME
VOGL \- A very ordinary GL Like Library.
.SH DESCRIPTION
.LP
.I VOGL
is a library of C routines which try to allow a programmer to write programs
which can be moved to machines which have the Silicon Graphics GL library
on them. It is based entirely on the VOGLE graphics library, and as
a result can handle circles, curves, arcs, patches, and polygons
in a device independent fashion. Simple hidden line removal is also
available via polygon backfacing. Access to hardware text and double
buffering of drawings depends on the driver. There is also a FORTRAN
interface but as it goes through the C routines FORTRAN users are warned
that arrays are in row-column order in C. Both the long FORTRAN names
and the shortened six character names are supported. People interested
in using software text should see the hershey library,
.I HERSHEY(3).
Some routines are only available in VOGL. If you include them in programs it
is advisable to put #ifdef VOGL ... #endif around them. The constant VOGL
is defined whenever a VOGL header file is included.
It should be noted that there are a number of routines that take the
type
.I Angle
for some of their parameters. All angles specified this way are actually
.I Integer Tenths Of Degrees.
(Don't ask!)
.SS Include files.
.LP
There are two include files provided with vogl: vogl.h and vodevice.h.
The file vogl.h has the type definitions and function interfaces, ideally
it is included where you would include gl.h on an SGI. The file vodevice.h
has the devices in it, and it is included where you would include device.h
on an SGI.
.SS
The following is a brief summary of the VOGL subroutines.
.SS Using X toolkits ans Sunview
.LP
For X11 and Sunview based applications, it is posible for VOGL to use a window that is supplied by that application's
toolkit. Under these circumstances, the toolkit is is responsible for handling
of all input events, and VOGL simply draws into the supplied window.
These calls are only available from C. Also see the directories examples/xt,
examples/xview and examples/sunview.
For X based toolkits the following two calls may be used:
.TP
.I vo_xt_window(display, xwin, width, height)
Tells VOGL to use the supplied window
.IB xwin
.nf
vo_xt_window(display, xwin, width, height)
Display *display;
Window xwin;
int width, height;
.fi
This routine should be called before calling "ginit()".
.TP
.I vo_xt_win_size(width, height)
Tells VOGL that the supplied window has changed size.
.nf
vo_xt_win_size(width, height)
int width, height;
.fi
.TP
For sunview based applications the following two calls may be used:
.TP
.I vo_sunview_canvas(canvas, width, height)
Tells VOGL to use the supplied sunview canvas
.IB canvas
.TP
.nf
vo_sunview_canvas(canvas, width, height)
Canvas canvas;
int width, height;
.fi
This routine should be called before calling "ginit()".
.TP
.I vo_sunview_canvas_size(width, height)
Tells VOGL that the supplied canvas has changed size.
.nf
vo_sunview_canvas_size(width, height)
int width, height;
.fi
.SS Device routines.
.TP
.I vinit(device)
Tell VOGL what the device is. This routine needs to be called if
the environment variable VDEVICE isn't set, or if the value in VDEVICE
is not to be used.
.nf
Fortran:
subroutine vinit(device, len)
character *(*) device
integer len
C:
vinit(device);
char *device;
Note 1 :- Current available devices are:
tek - tektronix 4010 and compatibles
hpgl - HP Graphics language and compatibles
dxy - roland DXY plotter language
postscript - monochrome postscript devices
ppostscript - monochrome postscript devices (portrait mode)
cps - colour postscript devices
pcps - colour postscript devices (portrait mode)
grx - the GRX library that is part of DJGPP.
- (little tested)
sun - Sun workstations running sunview
X11 - X windows (SUN's Openwindows etc etc)
decX11 - the decstation (old) window manager
This is only included in case you need it.
apollo - Apollo workstations
NeXT - NeXTStep
hercules - IBM PC hercules graphics card
cga - IBM PC cga graphics card
ega - IBM PC ega graphics card
vga - IBM PC vga graphics card
sigma - IBM PC sigma graphics card.
mswin - MS-windoze (little tested).
Sun, X11, decX11, apollo, hercules, cga
and ega support double buffering.
Note 2 :- If device is a NULL or a null string the value
of the environment variable "VDEVICE" is taken as the
device type to be opened.
Note 3 :- after init it is wise to explicitly
clear the screen.
e.g.: in C
color(BLACK);
clear();
or in Fortran
call color(BLACK)
call clear
.fi
.TP
.I ginit()
Open the graphics device and do the basic initialisation. This routine
is marked for obsolescence. The routine
.I winopen
(see below) should be used instead.
Note: this automatically causes a REDRAW event to appear as the first event
in the event queue.
.nf
Fortran:
subroutine ginit
C:
ginit()
.fi
.TP
.I winopen(title)
Open the graphics device and do the basic initialisation. This routine
should be used instead of
.I ginit.
Note: this automatically causes a REDRAW event to appear as the first event
in the event queue.
.nf
Fortran:
subroutine winopen(title, len)
character*(*) title
integer len
C:
winopen(title)
char *title;
.fi
.TP
.I gexit()
Reset the window/terminal (must be the last VOGL routine called)
.nf
Fortran:
subroutine gexit
C:
gexit()
.fi
.TP
.I voutput(path)
Redirect output from *next* ginit to file given by path. This routine only
applies to devices drivers that write to stdout e.g. postscript and hpgl.
.nf
Fortran:
subroutine voutput(path, len)
character*(*) path
integer len
C:
voutput(path)
char *path;
.fi
.TP
.I vnewdev(device)
Reinitialize VOGL to use a new device without changing attributes, viewport
etc.
(eg. window and viewport specifications)
.nf
Fortran:
subroutine vnewdev(device, len)
character *(*) device
integer len
C:
vnewdev(device)
char *device;
.fi
.I getplanes()
Returns the number of bit planes (or color planes) for a particular
device. The number of colors displayable by the device is then 2**(nplanes-1)
.nf
Fortran:
integer function getplanes()
C:
long
getplanes()
.fi
.SS Routines for controling flushing or syncronisation of the display.
On some devices (particularly X11) considerable speedups in display
can be achieved by not flushing each graphics primitive call to the
actual display until necessary. VOGL automatically delays flushing
under in following cases:
.nf
- Within a callobj() call.
- Within curves and patches.
- Within bgn*/end* calls.
- When double buffering (the flush is only done withing swapbuffers).
.fi
There are two user routines (which are NOT GL compatible) that can be used
to control flushing.
.TP
.I vsetflush(yesno)
Set global flushing status. If yesno = 0 (.false.) then don't do any
flushing (except in swapbuffers(), or vflush()). If yesno = 1 (.true.)
then do the flushing as described above.
.nf
Fortran:
subroutine vsetflush(yesno)
logical yesno
C:
void
vsetflush(yesno)
int yesno;
.fi
.TP
.I vflush()
Call the device flush or syncronisation routine. This forces a flush.
.nf
Fortran:
subroutine vflush
C:
void
vflush();
.fi
.SS Routines For Setting Up Windows.
Some devices are basically window orientated - like sunview and X11. You
can give VOGL some information about the window that it will use with these
routines.
These can make your code very device dependent. Both routines take
arguments which are in device space. (0, 0) is the bottom left hand corner
in device space. To have any effect these routines must be called before ginit
or winopen.
For the X11 device, an entry may be made in your .Xdefaults file of the
form vogl.Geometry =150x500+550+50 (where you specify your geometry as
you please).
.TP
.I prefposition(x1, x2, y1, y2)
Specify the preferred position of the window opened by the *next* winopen.
.nf
Fortran:
subroutine prefposition(x1, x2, y1, y2)
integer x1, x2, y1, y2
C:
prefposition(x1, x2, y1, y2)
long x1, x2, y1, y2
.fi
.TP
.I prefsize(width, height)
Specify the preferred width and height of the window opened by the
*next* winopen.
.nf
Fortran:
subroutine prefsize(width, height)
integer width, height
C:
prefsize(width, height)
long width, height;
.fi
.TP
.I reshapeviewport
This is occasionally used in Iris GL if a REDRAW event rolls up. While
VOGL is unlikely to ever provide a REDRAW event (except possibly the first event
in the event queue) the call is provided for
compatibility.
.nf
Fortran:
subroutine reshap
C:
reshapeviewport()
.fi
.SS General Routines.
.TP
.I clear()
Clears the current viewport to the current colour.
.nf
Fortran:
subroutine clear
C:
clear()
.fi
.TP
.I color(col)
Set the current colour. The standard colours are as follows:
.nf
black = 0 red = 1 green = 2 yellow = 3
blue = 4 magenta = 5 cyan = 6 white = 7.
These are included in vogl.h as:
BLACK, RED, GREEN, YELLOW, BLUE, MAGENTA, CYAN and WHITE.
When using fortran these are included in fvogl.h as
BLACK, RED, GREEN, YELLOW, BLUE, MAGENT, CYAN and WHITE.
.fi
.nf
Fortran:
subroutine color(col)
integer col
C:
color(col)
Colorindex col;
.fi
.TP
.I colorf(col)
Same as
.I color
only it takes a floating point argument. In Iris GL there
are sometimes good reasons for using this routine over
.I color.
See the GL manual for more details.
.nf
Fortran:
subroutine colorf(col)
real col
C:
colorf(col)
float col;
.fi
.TP
.I mapcolor(indx, red, green, blue)
Set the color map index indx to the color represented by (red, green, blue).
If the device has no color map this call does nothing.
.nf
Fortran:
subroutine mapcolor(indx, red, green, blue)
integer indx, red, green, blue
C:
mapcolor(indx, red, green, blue)
Colorindex indx;
short red, green, blue;
.fi
.TP
.I defbasis(id, mat)
Define basis number id to be the matrix mat.
.nf
Fortran:
subroutine defbasis(id, mat)
integer id
real mat(4, 4)
C:
defbasis(id, mat)
short id;
Matrix mat;
.fi
.TP
.I polymode(mode)
.I NOTE:- For this call to
.I have any effect
it must have been
.I conditionally compilied
into the library. (See polygons.c for details)
Control the filling of polygons. It expects one of the following
PYM_LINE, which means only the edges of the polygon will be drawn
and PYM_FILL which means fill the polygon (the default). PYM_POINT
and PYM_HOLLOW
are also recognised but they don't behave quite as they would
with SGI GL.
Also note that in Fortran the corresponding constants are truncated to
PYM_LI, PYM_FI, PYM_PO and PYM_HO respectivly. These appear in fvogl.h.
.nf
Fortran:
subroutine polymode(mode)
integer mode
C:
polymode(mode)
long mode;
.fi
.SS The Device Queue and Valuator Routines.
The available devices are defined in the header files vodevice.h and
for FORTRAN fvodevice.h
.TP
.I qdevice(dev)
Enable a device. Note: in VOGL the queue is of length 1.
.nf
Fortran:
subroutine qdevice(dev)
integer dev
C:
qdevice(dev)
Device dev;
.fi
.TP
.I unqdevice(dev)
Disable a device.
.nf
Fortran:
subroutine qdevice(dev)
integer dev
C:
qdevice(dev)
Device dev;
.fi
.TP
.I qread(data)
Read an event from the device queue. This routines blocks until
something happens. Note: it is important to have called qdevice
before doing this.
.nf
Fortran:
integer function qread(data)
integer*2 data
C:
long qread(data)
short *data;
.fi
.TP
.I isqueued(dev)
Check to see if device dev is enabled for queueing.
.nf
Fortran:
logical function isqueued(dev)
integer dev
C:
Boolean isqueued(dev)
short *dev;
.fi
.TP
.I qtest()
Check if there is anything in the queue. Note: in VOGL the queue
is only 1 entry deep.
.nf
Fortran:
logical function qtest
C:
Boolean qtest()
.fi
.TP
.I qreset()
Reset the device queue. This will get rid of any pending events.
.nf
Fortran:
subroutine qreset
C:
qreset()
.fi
.TP
.I getbutton(dev)
Returns the up (0) or down (1) state of a button.
.nf
Fortran:
logical function getbutton(dev)
integer dev
C:
Boolean getbutton(dev)
Device dev;
.fi
.TP
.I getvaluator(dev)
Return the current value of the valuator. Currently the only
valuators supported are MOUSEX and MOUSEY.
.nf
Fortran:
integer function getvaluator(dev)
integer dev
C:
long getvaluator(dev)
Device dev;
.fi
.SS Viewport Routines.
.TP
.I viewport(left, right, bottom, top)
Specify which part of the screen to draw in. Left, right, bottom, and top
are integer values in screen coordinates.
.nf
Fortran:
subroutine viewport(left, right, bottom, top)
integer left, right, bottom, top
C:
viewport(left, right, bottom, top)
Screencoord left, right, bottom, top;
.fi
.TP
.I pushviewport()
Save current viewport on the viewport stack.
.nf
Fortran:
subroutine pushviewport
C:
pushviewport()
.fi
.TP
.I popviewport()
Retrieve last pushed viewport.
.nf
Fortran:
subroutine popviewport
C:
popviewport()
.fi
.TP
.I getviewport(left, right, bottom, top)
Returns the left, right, bottom and top limits of the current viewport
in screen coordinates.
.nf
Fortran:
subroutine getviewport(left, right, bottom, top)
integer*2 left, right, bottom, top
C:
getviewport(left, right, bottom, top)
Screencoord *left, *right, *bottom, *top;
.fi
.SS Attribute Stack Routines.
.LP
The attribute stack contains details such as current color, current line style
and width, and the current font number. If you
need to prevent object calls form changing these, use
.I pushattributes
before the call and
.I popattributes
after.
.TP
.I pushattributes()
Save the current attributes on the attribute stack.
.nf
Fortran:
subroutine pushattributes
C:
pushattributes()
.fi
.TP
.I popattributes()
Restore the attributes to what they were at the last
.I pushattribute().
.nf
Fortran:
subroutine popattributes
C:
popattributes()
.fi
.SS Projection Routines.
.LP
All the projection routines define a new transformation matrix, and
consequently the world units. Parallel projections are defined by ortho or
ortho2. Perspective projections can be defined by perspective and window.
Note the types Angle, etc, are defined in vogl.h. Remember angles are in tenths
of degrees.
.TP
.I ortho(left, right, bottom, top, near, far)
Define x (left, right), y (bottom, top), and z (near, far) clipping
planes. The near and far clipping planes are actually specified as
distances along the line of sight. These distances can also be negative.
The actual location of the clipping planes is z = -near_d and z = -far_d.
.nf
Fortran:
subroutine ortho(left, right, bottom, top, near_d, far_d)
real left, right, bottom, top, near_d, far_d
C:
ortho(left, right, bottom, top, near_d, far_d)
Coord left, right, bottom, top, near_d, far_d;
.fi
.TP
.I ortho2(left, right, bottom, top)
Define x (left, right), and y (bottom, top) clipping planes.
.nf
Fortran:
subroutine ortho2(left, right, bottom, top)
real left, right, bottom, top
C:
ortho2(left, right, bottom, top)
float left, right, bottom, top;
.fi
.TP
.I perspective(fov, aspect, near, far)
Specify a perspective viewing pyramid in world coordinates by
giving a field of view, aspect ratio and the distance from the
eye of the near and far clipping plane.
.nf
Fortran:
subroutine perspective(fov, aspect, near, far)
integer fov
real aspect, near, far
C:
perspective(fov, aspect, near, far)
Angle fov;
float aspect;
Coord near, far;
.nf
.TP
.I window(left, right, bot, top, near, far)
Specify a perspective viewing pyramid in world coordinates by
giving the rectangle closest to the eye (ie. at the near clipping
plane) and the distances to the near and far clipping planes.
.nf
Fortran:
subroutine window(left, right, bot, top, near, far)
real left, right, bot, top, near, far
C:
window(left, right, bot, top, near, far)
float left, right, bot, top, near, far;
.fi
.SS Matrix Stack Routines.
.TP
.I pushmatrix()
Save the current transformation matrix on the matrix stack.
.nf
Fortran:
subroutine pushmatrix
C:
pushmatrix()
.fi
.TP
.I popmatrix()
Retrieve the last matrix pushed and make it the current transformation
matrix.
.nf
Fortran:
subroutine popmatrix
C:
popmatrix()
.fi
.SS Viewpoint Routines.
.LP
Viewpoint routines alter the current tranformation matrix.
.TP
.I polarview(dist, azim, inc, twist)
Specify the viewer's position in polar coordinates by giving
the distance from the viewpoint to the world origin,
the azimuthal angle in the x-y plane, measured from the y-axis,
the incidence angle in the y-z plane, measured from the z-axis,
and the twist angle about the line of sight.
.nf
Fortran:
subroutine polarview(dist, azim, inc, twist)
real dist
integer azim, inc, twist
C:
polarview(dist, azim, inc, twist)
Coord dist;
Angle azim, inc, twist;
.fi
.TP
.I
lookat(vx, vy, vz, px, py, pz, twist)
Specify the viewer's position by giving a viewpoint and a
reference point in world coordinates. A twist about the line
of sight may also be given.
.nf
Fortran:
subroutine lookat(vx, vy, vz, px, py, pz, twist)
real vx, vy, vz, px, py, pz
integer twist
C:
lookat(vx, vy, vz, px, py, pz, twist)
float vx, vy, vz, px, py, pz;
Angle twist;
.fi
.SS Move Routines.
.LP
There are variations on all these routines that end in 's' and also
end in 'i'. In the case of the 's' variations they take arguments
of type Scoord in C and integer*2 in FORTRAN. In the case of the 'i'
variations they take arguments of type Icoord in C and integer in
FORTRAN.
.TP
.I move(x, y, z)
Move current graphics position to (x, y, z). (x, y, z) is a point in
world coordinates.
.nf
Fortran:
subroutine move(x, y, z)
real x, y, z
C:
move(x, y, z)
Coord x, y, z;
.fi
.TP
.I rmv(deltax, deltay, deltaz)
Relative move. deltax, deltay, and deltaz are offsets in world
units.
.nf
Fortran:
subroutine rmv(deltax, deltay, deltaz)
real deltax, deltay, deltaz
C:
rmv(deltax, deltay, deltaz)
Coord deltax, deltay, deltaz;
.fi
.TP
.I move2(x, y)
Move graphics position to point (x, y). (x, y) is a point in world
coordinates.
.nf
Fortran:
subroutine move2(x, y)
real x, y
C:
move2(x, y)
Coord x, y;
.fi
.TP
.I rmv2(deltax, deltay)
Relative move2. deltax and deltay are offsets in world units.
.nf
Fortran:
subroutine rmv2(deltax, deltay)
real deltax, deltay
C:
rmv2(deltax, deltay)
Coord deltax, deltay;
.fi
.SS Line routines.
.LP
These routines set the line style and line width if the current device
is capable of doing so.
.TP
.I deflinestyle(n, style)
Define a line style and binds it to the integer n. The line style is a
bit pattern of 16 bits width.
.nf
Fortran:
subroutine deflin(n, style)
integer n
integer style
C:
deflinestyle(n, style)
short n;
Linestyle style;
.fi
.TP
.I setlinestyle(n)
Sets the current line style.
.nf
Fortran:
subroutine setlin(n)
integer n
C:
setlinestyle(n)
short n;
.fi
.TP
.I linewidth(n)
Sets the current line width to 'n' pixels wide.
.nf
Fortran:
subroutine linewi(n)
integer n
C:
linewidth(n)
short n;
.fi
.SS Drawing Routines.
.LP
There are variations on all these routines that end in 's' and also
end in 'i'. In the case of the 's' variations they take arguments
of type Scoord in C and integer*2 in FORTRAN. In the case of the 'i'
variations they take arguments of type Icoord in C and integer in
FORTRAN.
.TP
.I draw(x, y, z)
Draw from current graphics position to (x, y, z). (x, y, z) is a point in
world coordinates.
.nf
Fortran:
subroutine draw(x, y, z)
real x, y, z
C:
draw(x, y, z)
Coord x, y, z;
.fi
.TP
.I rdr(deltax, deltay, deltaz)
Relative draw. deltax, deltay, and deltaz are offsets in world units.
.nf
Fortran:
subroutine rdr(deltax, deltay, deltaz)
real deltax, deltay, deltaz
C:
rdr(deltax, deltay, deltaz)
Coord deltax, deltay, deltaz;
.fi
.TP
.I draw2(x, y)
Draw from current graphics position to point (x, y). (x, y) is a point in
world coordinates.
.nf
Fortran:
subroutine draw2(x, y)
real x, y
C:
draw2(x, y)
Coord x, y;
.fi
.TP
.I rdr2(deltax, deltay)
Relative draw2. deltax and deltay are offsets in world units.
.nf
Fortran:
subroutine rdr2(deltax, deltay)
real deltax, deltay
C:
rdr2(deltax, deltay)
Coord deltax, deltay;
.fi
.SS Vertex calls.
.LP
There are calls which we term 'vertex calls' which simply specify a point
in 4D, 3D or 2D. These calls take an array which specifies the coordinates
of the point. The interpretation of these points is described below.
.I v4d(v)
Specify a vertex(point) in 4D using double precision numbers.
.nf
Fortran:
subroutine v4d(v)
real *8 v(4)
C:
v4d(v)
double v[4];
.fi
.I v4f(v)
Specify a vertex(point) in 4D using single precision floating point numbers.
.nf
Fortran:
subroutine v4f(v)
real v(4)
C:
v4f(v)
float v[4];
.fi
.I v4i(v)
Specify a vertex(point) in 4D using integer numbers
.nf
Fortran:
subroutine v4i(v)
integer v(4)
C:
v4i(v)
long v[4];
.fi
.I v4s(v)
Specify a vertex(point) in 4D using short integer numbers
.nf
Fortran:
subroutine v4s(v)
integer *2 v(4)
C:
v4s(v)
short v[4];
.fi
.LP
There are also equivalent calls for 3D points (v3d, v3f, v3i, v3s)
and 2D points (v2d, v2f, v2i, v2s). The only difference is the number
of elements that each vertex needs to be specified. It should also be
noted the the different data types (ie. double, float, long and short)
are merely different ways of representing the same basic coordinate
data (calling v3s with v[] = {100,200,200} is the same as calling v3f
with v[] = {100.0, 200.0, 200.0}).
The way these points are interpreted depends on what mode
has be set up with one of the calls
.I bgnpoint, bgnline, bgnclosedline or bgnpolygon.
The
.I bgnpoint
call specifies that the next series of vertex calls are specifying a chain
of points (dots) to be drawn. A
.I bgnpoint
is terminated with a
.I endpoint
call.
.nf
Fortran:
subroutine bgnpoint
C:
bgnpoint()
Fortran:
subroutine endpoint
C:
endpoint()
.fi
The
.I bgnline
call specifies that the next series of vertex calls are specifying the points
on a polyline. A
.I bgnline
is terminated with a
.I endline
call.
.nf
Fortran:
subroutine bgnline
C:
bgnline()
Fortran:
subroutine endline
C:
endline()
.fi
The
.I bgnclosedline
call is similar to the
.I bgnline
except that when
.I endclosedline
is called the first point given (ie. the one first after the bgnclosedline
call) is joined to the last point given (ie. the one just before the
endclosedline call).
.nf
Fortran:
subroutine bgncloseline
C:
bgnclosedline()
Fortran:
subroutine endclosedline
C:
endclosedline()
.fi
The
.I bgnpolygon
call specifies that the next series of vertex calls are defining a polygon.
When
.I endpolygon
is called, the polygon is closed and filled (or drawn as an outline depending
on the mode that has been set with the
.I polymode
call if this call has been compilied into the library.
.nf
Fortran:
subroutine bgnpolygon
C:
bgnpolygon()
Fortran:
subroutine endpolygon
C:
endpolygon()
.fi
.SS Arcs and Circles.
.LP
There are variations on all these routines that end in 's' and also
end in 'i'. In the case of the 's' variations they take arguments
of type Scoord in C and integer*2 in FORTRAN. In the case of the 'i'
variations they take arguments of type Icoord in C and integer in
FORTRAN.
.TP
.I circleprecision(nsegs)
Set the number of line segments making up a circle. Default is
currently 32. The number of segments in an arc is
calculated from nsegs according the span of the arc.
This routine is only available in VOGL.
.nf
Fortran:
subroutine circleprecision(nsegs)
integer nsegs
C:
circleprecision(nsegs)
int nsegs;
.fi
.TP
.I arc(x, y, radius, startang, endang)
Draw an arc. x, y, and radius are values in world units.
.nf
Fortran:
subroutine arc(x, y, radius, startang, endang)
real x, y, radius;
integer startang, endang;
C:
arc(x, y, radius, startang, endang)
Coord x, y, radius;
Angle startang, endang;
.fi
.TP
.I arcf(x, y, radius, startang, endang)
Draw a filled arc. x, y, and radius are values in world units. (How
the filling is done may be changed by calling
.I polymode
, if this call has been compilied into the library).
.nf
Fortran:
subroutine arcf(x, y, radius, startang, endang)
real x, y, radius;
integer startang, endang;
C:
arcf(x, y, radius, startang, endang)
Coord x, y, radius;
Angle startang, endang;
.fi
.TP
.I circ(x, y, radius)
Draw a circle. x, y, and radius are values in world units.
.nf
Fortran:
subroutine circ(x, y, radius)
real x, y, radius
C:
circ(x, y, radius)
Coord x, y, radius;
.fi
.TP
.I circf(x, y, radius)
Draw a filled circle. x, y, and radius are values in world units. How
the filling is done may be changed by calling
.I polymode.
.nf
Fortran:
subroutine circf(x, y, radius)
real x, y, radius
C:
circf(x, y, radius)
Coord x, y, radius;
.fi
.SS Curve Routines.
.TP
.I curvebasis(id)
Set the basis matrix for a curve to the matrix referenced by id.
The matrix and it's id are tied together with a call to
.I defbasis.
.nf
Fortran:
subroutine curvebasis(id)
integer id
C:
curvebasis(id)
short id;
.fi
.TP
.I curveprecision(nsegs)
Define the number of line segments used to draw a curve.
.nf
Fortran:
subroutine curveprecision(nsegs)
integer nsegs
C:
curveprecision(nsegs)
short nsegs;
.fi
.TP
.I rcrv(geom)
Draw a rational curve.
.nf
Fortran:
subroutine rcrv(geom)
real geom(4,4)
C:
rcrv(geom)
Coord geom[4][4];
.fi
.TP
.I rcrvn(n, geom)
Draw n - 3 rational curve segments. Note: n must be at least 4.
.nf
Fortran:
subroutine rcrvn(n, geom)
integer n
real geom(4,n)
C:
rcrvn(n, geom)
long n;
Coord geom[][4];
.fi
.TP
.I crv(geom)
Draw a curve.
.nf
Fortran:
subroutine crv(geom)
real geom(3,4)
C:
crv(geom)
Coord geom[4][3];
.fi
.TP
.I crvn(n, geom)
Draw n - 3 curve segments. Note: n must be at least 4.
.nf
Fortran:
subroutine crvn(n, geom)
integer n
real geom(3,n)
C:
crvn(n, geom)
long n;
Coord geom[][3];
.fi
.TP
.I curveit(n)
Draw a curve segment by iterating the top matrix in the matrix stack as
a forward difference matrix. This performs 'n' iterations.
.nf
Fortran:
subroutine curveit(n)
integer n
C:
curveit(n)
short n;
.fi
.SS Rectangles and General Polygon Routines.
.LP
See also
.I Vertex
calls above.
The way in which filled polygons (including circles and arcs) are
treated depends on the mode that has been set with the
.I polymode
call.
.LP
There are variations on all these routines that end in 's' and also
end in 'i'. In the case of the 's' variations they take arguments
of type Scoord in C and integer*2 in FORTRAN. In the case of the 'i'
variations they take arguments of type Icoord in C and integer in
FORTRAN.
.TP
.I rect(x1, y1, x2, y2)
Draw a rectangle.
.nf
Fortran:
subroutine rect(x1, y1, x2, y2)
real x1, y1, x1, y2
C:
rect(x1, y1, x2, y2)
Coord x1, y1, x2, y2;
.fi
.TP
.I rectf(x1, y1, x2, y2)
Draw a filled rectangle. (How the filling is done may be changed by calling
.I polymode
, if this call has been compilied into the library).
.nf
Fortran:
subroutine rectf(x1, y1, x2, y2)
real x1, y1, x1, y2
C:
rectf(x1, y1, x2, y2)
Coord x1, y1, x2, y2;
.fi
.TP
.I poly2(n, points)
Construct a (x, y) polygon from an array of points provided by the user.
.nf
Fortran:
subroutine poly2(n, points)
integer n
real points(2, n)
C:
poly2(n, points)
long n;
Coord points[][2];
.fi
.TP
.I polf2(n, points)
Construct a filled (x, y) polygon from an array of points provided by the user.
(How the filling is done may be changed by calling
.I polymode
, if this call has been compilied into the library).
.nf
Fortran:
subroutine polf2(n, points)
integer n
real points(2, n)
C:
polf2(n, points)
long n;
Coord points[][2];
.fi
.TP
.I poly(n, points)
Construct a polygon from an array of points provided by the user.
.nf
Fortran:
subroutine poly(n, points)
integer n
real points(3, n)
C:
poly(n, points)
long n;
float points[][3];
.fi
.TP
.I polf(n, points)
Construct a filled polygon from an array of points provided by the user.
(How the filling is done may be changed by calling
.I polymode
, if this call has been compilied into the library).
.nf
Fortran:
subroutine polf(n, points)
integer n
real points(3, n)
C:
polf(n, points)
long n;
Coord points[][3];
.fi
.TP
.I backface(onoff)
Turns on culling of backfacing polygons. A polygon is
backfacing if it's orientation in *screen* coords is clockwise.
.nf
Fortran:
subroutine backface(onoff)
logical onoff
C:
backface(onoff)
Boolean onoff;
.fi
.TP
.I frontface(onoff)
Turns on culling of frontfacing polygons. A polygon is
frontfacing if it's orientation in *screen* coords is anticlockwise.
.nf
Fortran:
subroutine frontface(clockwise)
logical onoff
C:
frontface(clockwise)
Boolean onoff;
.fi
.SS Text routines.
The original VOGLE hardware fonts "small" and "large" have the font numbers
0 and 1 respectively. The default font is 0. For X11 displays the default
fonts used by the program can be overridden by placing the following defaults
in the ~/.Xdefaults file:
.nf
vogl.smallfont: <font name>
vogl.largefont: <font name>
.fi
.TP
.I font(fontid)
Set the current font
.nf
Fortran:
subroutine font(fontid)
integer fontid;
C:
font(fontid)
short fontid;
.fi
.TP
.I cmov(x, y, z)
Change the current character position. The usual variations with the
extensions 'i' and 's' also apply here.
.nf
Fortran:
subroutine cmov(x, y, z)
real x, y, z;
C:
cmov(x, y, z)
Coord x, y, z;
.fi
.TP
.I cmov2(x, y)
Change the current character position in x and y. The usual variations
with the extensions 'i' and 's' also apply here.
.nf
Fortran:
subroutine cmov2(x, y)
real x, y;
C:
cmov2(x, y)
Coord x, y;
.fi
.TP
.I getheight()
Return the maximum height in the current font.
.nf
Fortran:
integer function getheight
C:
long
getheight()
.fi
.TP
.I strwidth(s)
Return the length of the string s in screen coords.
.nf
Fortran:
integer function strwidth(s, n)
character *(*) s
integer n;
C:
long
strwidth(s)
char *s;
.fi
.TP
.TP
.I charstr(str)
Draw the text in string at the current position.
.nf
Fortran:
subroutine charst(str, len)
character*(*) str
integer len
C:
charstr(str)
char *str;
.fi
.SS Transformations Routines.
.LP
All transformations are cumulative, so if you rotate something and then
do a translate you are translating relative to the rotated axes. If you need
to preserve the current transformation matrix use pushmatrix(), do the
drawing, and then call popmatrix() to get back where you were before.
.TP
.I translate(x, y, z)
Set up a translation.
.nf
Fortran:
subroutine translate(x, y, z)
real x, y, z
C:
translate(x, y, z)
Coord x, y, z;
.fi
.TP
.I scale(x, y, z)
Set up scaling factors in x, y, and z axis.
.nf
Fortran:
subroutine scale(x, y, z)
real x, y, z
C:
scale(x, y, z)
Coord x, y, z;
.fi
.TP
.I rot(angle, axis)
Set up a rotation in axis axis. Axis is one of 'x', 'y', or 'z'.
The angle in this case is a real number in degrees.
.nf
Fortran:
subroutine rot(angle, axis)
real angle
character axis
C:
rot(angle, axis)
float angle;
char axis;
.fi
.TP
.I rotate(angle, axis)
Set up a rotation in axis axis. Axis is one of 'x', 'y', or 'z', and
the angle is in tenths of degrees. Makes you feel sentimental doesn't it.
.nf
Fortran:
subroutine rotate(angle, axis)
integer angle
character axis
C:
rotate(angle, axis)
Angle angle;
char axis;
.fi
.SS Patch Routines.
.TP
.I patchbasis(tbasisid, ubasisid)
Define the t and u basis matrix id's of a patch. It is assumed that tbasisid
and ubasisid have matrices associated with them already (this is done using
the
.I defbasis
call).
.nf
Fortran:
subroutine patchbasis(tid, uid)
integer tid, uid
C:
patchbasis(tid, ubid)
long tid, uid
.fi
.TP
.I patchprecision(tseg, useg)
Set the minimum number of line segments making up curves in a patch.
.nf
Fortran:
subroutine patchprecision(tseg, useg)
integer tseg, useg
C:
patchprecision(tseg, useg)
long tseg, useg;
.fi
.TP
.I patchcurves(nt, nu)
Set the number of curves making up a patch.
.nf
Fortran:
subroutine patchcurves(nt, nu)
integer nt, nu
C:
patchcurves(nt, nu)
long nt, nu;
.fi
.TP
.I rpatch(gx, gy, gz, gw)
Draws a rational patch in the current basis, according to the geometry
matrices gx, gy, gz, and gw.
.nf
Fortran:
subroutine rpatch(gx, gy, gz, gw)
real gx(4,4), gy(4,4), gz(4,4), gw(4,4)
C:
rpatch(gx, gy, gz, gw)
Matrix gx, gy, gz, gw;
.fi
.TP
.I patch(gx, gy, gz)
Draws a patch in the current basis, according to the geometry
matrices gx, gy, and gz.
.nf
Fortran:
subroutine patch(gx, gy, gz)
real gx(4,4), gy(4,4), gz(4,4)
C:
patch(gx, gy, gz)
Matrix gx, gy, gz;
.fi
.SS Point Routines.
.LP
There are variations on all these routines that end in 's' and also
end in 'i'. In the case of the 's' variations they take arguments
of type Scoord in C and integer*2 in FORTRAN. In the case of the 'i'
variations they take arguments of type Icoord in C and integer in
FORTRAN.
.TP
.I pnt(x, y, z)
Draw a point at x, y, z
.nf
Fortran:
subroutine pnt(x, y, z)
real x, y, z
C:
pnt(x, y, z)
Coord x, y, z;
.fi
.TP
.I pnt2(x, y)
Draw a point at x, y.
.nf
Fortran:
subroutine pnt2(x, y)
real x, y
C:
pnt2(x, y)
Coord x, y;
.fi
.SS Object Routines.
.LP
Objects are graphical entities created by the drawing routines called between
.I makeobj
and
.I closeobj.
Objects may be called from within other objects. When an object
is created most of the calculations required by the drawing routines called
within it are done up to where the calculations involve the current
transformation matrix. So if you need to draw the same thing several times
on the screen but in different places it is faster to use objects than
to call the appropriate drawing routines each time.
.TP
.I makeobj(n)
Commence the object number n.
.nf
Fortran:
subroutine makeobj(n)
integer n
C:
makeobj(n)
Object n;
.fi
.TP
.I closeobj()
Close the current object.
.nf
Fortran:
subroutine closeobj()
C:
closeobj()
.fi
.TP
.I genobj()
Returns a unique object identifier.
.nf
Fortran:
integer function genobj()
C:
Object
genobj()
.fi
.TP
.I getopenobj()
Return the number of the current object.
.nf
Fortran:
integer function getopenobj()
C:
Object
getopenobj()
.fi
.TP
.I callobj(n)
Draw object number n.
.nf
Fortran:
subroutine callobj(n)
integer n
C:
callobj(n)
Object n;
.fi
.TP
.I isobj(n)
Returns non-zero if there is an object of number n.
.nf
Fortran:
logical function isobj(n)
integer n
C:
Boolean
isobj(n)
Object n;
.fi
.TP
.I delobj(n)
Delete the object number n.
.nf
Fortran:
subroutine delobj(n)
integer n
C:
delobj(n)
Object n;
.fi
.SS Double Buffering.
Where possible VOGL allows for front and back buffers
to enable things like animation and smooth updating of
the screen. Note: it isn't possible to have backbuffer and
frontbuffer true at the same time.
.TP
.I gconfig
With Iris GL you must call gconfig for things like doublebuffering
to take effect.
.nf
Fortran:
subroutine gconfig
C:
gconfig()
.fi
.TP
.I doublebuffer
Flags our intention to do double buffering.
.nf
Fortran:
subroutine doublebuffer
C:
doublebuffer()
.fi
.TP
.I singlebuffer
Switch back to singlebuffer mode.
.nf
Fortran:
subroutine singlebuffer
C:
singlebuffer()
.fi
.TP
.I backbuffer(Boolean)
Make VOGL draw in the backbuffer.
.nf
Fortran:
subroutine backbuffer(yesno)
logical yesno;
C:
backbuffer(yesno)
Boolean yesno;
.fi
.TP
.I frontbuffer(Boolean)
Make VOGL draw in the front buffer.
.nf
Fortran:
subroutine frontbuffer(yesno)
logical yesno;
C:
frontbuffer(yesno)
Boolean yesno;
.fi
.TP
.I swapbuffers()
Swap the front and back buffers.
.nf
Fortran:
subroutine swapbuffers
C:
swapbuffers()
.fi
.SS Position Routines.
.TP
.I getgpos(x, y, z, w)
Gets the current graphics position in world coords.
.nf
Fortran:
subroutine getgpos(x, y, z, w)
real x, y, z
C:
getgpos(x, y, z, w)
Coord *x, *y, *z, *w;
.fi
.TP
.I getcpos(ix, iy)
Gets the current character position in screen coords.
.nf
Fortran:
subroutine getcpo(ix, iy)
integer ix, iy
C:
getcpos(ix, iy)
Scoord *ix, *iy;
.fi
.SH BUGS
.LP
Double buffering isn't supported on all devices.
.LP
The yobbarays may be turned on or they may be turned off.
