Active-DVI |
Active-DVI is a viewer for DVI files that also recognizes a new class of \special’s targeted to presentations via laptop computers: various visual effects can easily be incorporated to the presentation, via a companion advi.sty LATEX package.Active-DVI is copyright ©2001–2011 INRIA and distributed under the Gnu Library General Public License —see the LGPL file included in the distribution.
This manual is available as sevral HTML files, one single HTML file, or in PDF format.
Active-DVI is based on Mldvi, written and distributed by Alexandre Miquel under the LGPL license1. This constitutes the core rendering engine. Active-DVI has then been developed by
Jun Furuse Didier Rémy, and Pierre Weis, with contributions by Didier Le Botlan, Roberto Di Cosmo, Xavier Leroy, Alexandre Miquel, and Alan Schmitt.
After successful installation of Active-DVI, don’t forget to have a look at example presentations provided with the distribution, in the directories:
In some cases, your Unix system provider will ensure an up-to-date port of the Active-DVI binaries; in such a case, use the recommended procedure to download, compile, and install the software.
In any other case, you have to install Active-DVI manually.
To compile Active-DVI, you must install the camlimages library first, including its native code version.
Hence, download and install camlimages first. After successful installation of this library, download and extract the Active-DVI source files, move to the new source directory of Active-DVI, and compile Active-DVI itself.
This is the “Manual installation in two steps” that has been packaged for you in one single step by the Active-DVI development kit.
Download and install the entire Active-DVI development kit (a.k.a. adk) source tree provided by the distribution: it contains Active-DVI sources and the camlimages library sources, plus a set of Makefiles to automate the installation procedure. Then, just type make in the directory where the Active-DVI Development Kit has been extracted from the archive. If everything goes well, it automatically compiles and installs the necessary libraries and Active-DVI for you. Some fine tuning may be necessary, if you are not running a standard Linux distribution, such as RedHat, Mandrake, or Debian.
If you have to install some other software to solve installation problems, you should type make reconfigure to re-perform the configuration process before launching make again.
See examples/slitex/simplistic/ for a simplistic talk example. There, the command \pause is used to make the presenter to stop while displaying the document. More involved examples can be found in the directory examples of the distribution.
Active-DVI may execute programs and commands embedded into the DVI file. Hence, when playing a DVI file from an untrusted source, you should run advi with the -safer option that inhibits the execution of embedded applications. This warning applies in particular if you choose Active-DVI as your default meta-mail previewer for the application/x-dvi mime-type.
The default safety option is the -ask option: it tells Active-DVI to ask the user each time it must launch an application. (Note that in such a case Active-DVI asks only once to launch a given application: it remembers your previous decisions concerning the command and acts accordingly for the rest of the presentation.)
The second safety option is the above mentioned -safer option: it completely inhibits the execution of embedded applications.
The last safety option is -exec: if you call advi -exec, advi automatically and silently launches all embedded applications (this is useful to play your own presentations without the burden of answering yes to Active-DVI’s questions).
As mentioned, the safe -ask option is the default, automatically set when nothing has been explicitly specified by the user. If desired, the default safety option can be set via initialization files, either on a system large scale by the machine administrator (in the file /etc/advirc), on a local scale by individual users (setting the default policy for that user), or even on a per directory basis (setting the default policy to show DVI files in this directory)! (This last option is convenient to gracefully run your own talks, while still being cautious when running talks from others.)
An initialization file for Active-DVI is simply a text file that contains options exactly similar to those you can give on the command line (with the exception of comments, made of a sharp sign (#) followed by some text that is ignored until the end of line). For instance:
-exec # I know what I mean! -bgcolor grey16 -fgcolor grey95
is a valid initialization file that sets the safety policy to -exec, then sets the background and foreground colors to obtain a nice reverse video effect.
Before parsing options on the command line, Active-DVI loads, in the order listed below, the following initialization files (nothing happens if any of them does not exist):
In addition, the user may load an arbitrary file containing options by specifying the file path via the command line argument -options-file. Hence, -options-file filename loads filename when parsing this option to set up the options contained in filename (thus overriding the options set before by the default ~/.advirc, ~/.advi/advirc, or ./.advirc, initialization files).
Active-DVI is invoked with the following command syntax
advi [options] dvifile [dvifile]
Once Active-DVI is launched, just press ? to get help
on the keys you can use to control the presenter (type ^f
(Control-F) to get full screen, < or > to change
the magnification of the text).
When two file names are provided Active-DVI displays them both: see section 5.12 for details on the use of a secondary DVI file.
The advi commands recognized the following options:
-v | Prints the advi current version and exits |
--version | Prints the full advi current version and exits |
-help | Short command line options help |
-geometry geom | Geometry of Active-DVI’s window specification |
Geometry geom is specified in pixels, using the standard format
for X-Window geometry specifications (i.e:
WIDTHxHEIGHT[+XOFFSET+YOFFSET]).
-fullwidth | Adjust the size of the window to the width of the screen |
-nomargins | Cancel horizontal and vertical margins |
-hmargin dimen | Horizontal margin specification (default: 1cm) |
-vmargin dimen | Vertical margin specification (default: 1cm) |
Dimensions are specified as numbers optionally followed by two letters representing units. When no units are given, dimensions are treated as numbers of pixels. Currently supported units are the standard TeX units as specified in the TeX book (D. Knuth, Addison-Wesley, (C) 1986): ‘pt’ (point), ‘pc’ (pica), ‘in’ (inch), ‘bp’ (big point), ‘cm’ (centimeter), ‘mm’ (millimeter), ‘dd’ (didot point), ‘cc’ (cicero) and ‘sp’ (scaled point). Note that dimensions are specified w.r.t the original TeX document, and do not correspond to what is actually shown on the screen, which can be displayed at a different resolution than specified in the original TeX source.
-crop | Crop the window to the best size (default) |
-nocrop | Disable cropping |
-fgcolor <color> | Specify the color of the foreground color |
-bgcolor <color> | Specify the color of the background color |
-rv | Specify that reverse video should be simulated by exchanging |
| the background and foreground colors | |
-gamma <float> | Specify gamma correction (>0.0) of glyphs |
-pager | Specify the name of the pager to launch on a txt link |
-browser | Specify the name of the browser to a html link |
--debug | General debug option |
--debug_pages | Debug page motion |
--show_ps | Print a copy of Postscript sent to gs to stdout |
--verbose_image_access | Change the cursor while loading images |
-A | Toggle Postscript anti-aliasing |
-passive | Inhibit effects that are visible when redrawing the page |
| (Transitions, delays, embedded applications) |
-exec | Set safety policy to “always execute embedded applications” |
-ask | Set safety policy to “ask user before execution of embedded applications” |
-safer | Set safety policy to “never execute embedded applications” |
-option-file <filename> | Load filename as a file containing a list of options |
| as given on the command line to advi. |
| -autoswitch | Set the autoswitch flag, which allows implicit switch to master |
| on usr1 signal (default is off). |
Text can be copied from the Active-DVI previewer to another application. However, this uses the XBuffer and not the XSelection mechanism.
Moreover, Shift left-click dump an ASCII representation of the region under the mouse pointer in the source file. This expects the DVI to have been instrumented with line numbers of the form
line: ⟨ line⟩ ⟨ file⟩
where ⟨ line⟩ and ⟨ file⟩ are the current source line and current source file.
The position is exported in ASCII, in the form
#line ⟨ before⟩, ⟨ after⟩
<<⟨ prefix⟩>><<⟨ suffix⟩>>
⟨ file⟩
where ⟨ before⟩ and ⟨ after⟩ are the enclosing line numbers, ⟨ prefix⟩⟨ suffix⟩ the word constituent surrounding the mouse position, and file is the name of the current file.
Line numbers default to 0 when not found. Note that line numbers may be inconsistent if there \special-line commands have not been inserted close enough to the position.
Active-DVI supports the LATEX hyperref package with both internal and cross-file references. For cross-file references, it launches a new advi process to view the target.
Active-DVI improves the treatment of hyper-refs over conventional previewers, by emphasizing the hyper-target text of an hyper-link. Thus, an hyper-target definition:
should make the activation of the link ⟨ text⟩ not only move to the page where ⟨ tag⟩ occurs, but also emphasize the destination target ⟨ tag⟩. However, since \hypertarget does include its second argument within the target, we use the following command instead:
\edef\hyper@quote{\string"}
\edef\hyper@sharp{\string#}
\newcommand{\softtarget}[2]%
{\special{html:<a name=\hyper@quote#1\hyper@quote>}#2%
\special{html:</a>}}
(If you are viewing this document with Active-DVI, you may move over this area or click on this one to see the effect.)
Similarly, to define a link target we use:
\newcommand{\softlink}[2]%
{\special{html:<a href=\hyper@quote\hyper@sharp#1\hyper@quote>}#2%
\special{html:</a>}}
There are two ways to include a floating table of contents while previewing.
t. The first stroke on t shows the first page
of the table of contents. Successive stokes will show the following
pages. (As usual, prefix integer argument may be used to directly
access a specific page of the table of contents.)The package advi described below redefines the macro \tableofcontents so that it automatically inserts the reserved hyper-targets markers around the table of contents. It also provides two new macros, \advitoc and \endadvitoc, that serve to insert these markers when the table of contents is hand-made.
Normally, thumbnails are drawn for all the pages. However, thumbnail pages can also be defined manually, with an hyper-target whose anchor is of the form /page.⟨ suffix⟩. In this case, all the desired thumbnails must be explicitly marked.
By default, the binding T processes thumbnails and the binding t displays thumbnails if already processed, or shows the table of contents if available. Otherwise pressing t has no action. Thumbnails computation is explicit, so that incidentally hitting the t key does not lead to an unexpected computation, hence an unexpected delay.
See the key bindings in the appendix.
During the show you can annotate your slides, entering the scratching mode. There are two modes, one for writing characters (the writing mode, entered by pressing s during the show) and the other to draw lines or figures on the slides (the drawing mode, entered by pressing S). In each of this modes, you can enter the scratch setting mode to set various properties of the scratching process. See the relevant key bindings for writing mode and drawing mode in the appendix.
If you press ^X-l (Control-X then l) the laser pointer appears
on the slide; the pointer sticks to the mouse pointer and allows easy
pointing to parts of the presentation. The laser pointer size and
color can be set on the command line (options -laser-pointer-color and laser-pointer-size).
You can save a snapshot of the current slide at any time by pressing
^X-^S (Control-X then Control-S). An image file is written
(by default a png file). The name of the file produced can be
set via the command line (see advi -help for details) or
directly from within the LATEX source file with commands
\advisavepageimage and \advisavepageimagefile{filename}).
Active-DVI provides the command \advipushkeys that provides
key presses to the presenter as if you had pressed it when viewing the
presentation. For instance:
\advipushkeys{"q"}
ends the talk immediately.
Note that control keys must be encoded inside key strings passed to
Active-DVI: we use the Emacs textual convention. For instance, the
character “Control-A” (ASCII 1) is denoted by the two charcters
^X (i.e. a carret character immediately followed by an X).
Hence, the command
\advipushkeys{"^X^F"}
switches to full screen mode.
Presentation examples can be found in the examples
directory. Don’t miss to play them! Then, feel free to read their
source code and copy the effects they provide.
Active-DVI can be used as is, but will shine when driven by a user with a bias towards programming: special effects can easily be realized by using the LATEX packages provided with the distribution.
Creative advanced users may program the presenter at various levels, either using or defining simple LATEX macros, writing new LATEX package files, or by implementing extensions to the previewer itself.
Active-DVI can be invoked with several DVI files (currently only two). The first file is always used as the master file and others are client files. The user can switch between files explicitly (see key bindings) or implicitly. There is an implicit switch from the master to the client file c when an hyperlink that is not found in the master file can be found in the client file c; there is also a switch from the client c to the master when using the history stack and the previous event on the stack was an implicit switch from the master to the client c.
If autoswitch flag is set, there is also an implicit switch to the master, whenever Active-DVI receives signal usr1 (to mean immediate refresh).
Active-DVI provides some LATEX packages to facilitate animations and interaction with the presenter from within your LATEX source text.
The advi.sty package is the main package to include when writing a presentation for Active-DVI. It defines the main set of interactive commands for Active-DVI to animate the show. However, there is no need to load the package if no Active-DVI special effects are required for the presentation.
Most commands of advi.sty use the TEX
\special command to insert into the DVI output file the
Active-DVI specific commands that implement their
semantics. Those commands are interpreted by Active-DVI afterwhile during
the DVI file previewing. Note that a \special{bla bla} command
is equivalent to a \hbox{} for TEX’s mouth, hence it may alter
the document layout accordingly. Thus, be aware that most commands of
the advi.sty package are equivalent to a \hbox{}
command as far as the document layout is concerned.
The advi.sty package recognizes the special option ignore, which helps the production of a printable version of the presentation: the ignore option makes the package not to produce Active-DVI specials, so that the show can be previewed by other DVI previewers or turned into Postscript using dvips. Of course, this option disables most effects that cannot be printed, although some of them are still approximated.
If the ignore option is not set globally, it can be set locally with the commands \adviignore. However, this will not prevent all effects, since some decisions are taken when the package is loaded.
The package also defines the conditional \ifadvi which evaluates its first argument if advi is not in ignore mode and its second argument otherwise.
Active-DVI provides partial display of pages (slide “strip-tease”): the Active-DVI’s rendering engine stops before the display of the current page is complete. The corresponding state is named a pause. Upon reaching a pause, Active-DVI may wait for a specified delay, or for user input. \adviwait[⟨ seconds⟩]
Wait for ⟨ seconds⟩. If no argument is provided, waits until the user requests to continue (hitting a key to move to next pause or to change page).
Active-DVI allows (almost) any piece of LATEX code to be recorded, and the corresponding DVI code to be rendered later upon request. To be able to render the code in any order we choose, we must bind the recorded LATEX code to a name (called a tag): this LATEX code together with its tag defines the notion of an Active-DVI record.
The entire DVI image of an Active-DVI record must fit on a single DVI page. The corresponding check is left to the writer of the document.
The command defining an Active-DVI record is as follows:
\advirecord[play]{⟨ tag⟩}{⟨ latex code⟩}
\begin{advirecording}[play]{⟨ tag⟩}{⟨ latex code⟩}⟨ text⟩\end{advirecording}
This command processes ⟨ latex code⟩ and records the corresponding DVI output, then binds it to the tag ⟨ tag⟩. While recording, the DVI output is not displayed, unless the option play is set.Active-DVI records may be nested. In this case, the inner record is bound to its own tag as usual; in addition, if the inner record is defined with the play option, it is also recorded as a part of the outer tag record.
If the environment syntax form of Active-DVI record definition is used, the ⟨ latex code⟩ may contain fragile commands.
To play an Active-DVI record, the corresponding DVI must have been recorded on the current DVI page and before issuing the play command. With the proviso that these requirements are satisfied, the syntax of the command to display an Active-DVI record is as follows: \adviplay[⟨ color⟩]{⟨ tag⟩}
This command plays the DVI code previously recorded and bound to ⟨ tag⟩.The optional argument changes the text color to ⟨ color⟩ during replay.
Active-DVI gives the ability to define Active-DVI anchors: an anchor is specified by (1) an Active-DVI record, (2) a LATEX piece of code that defines the area of the page where the anchor is active, and (3) an activation method.
The anchor is activated when some mouse events specified by the activation method occur in the area. The anchor is de-activated when the event that triggered the activation does not hold any more.
The Active-DVI record associated with the anchor is automatically rendered anew each time the anchor is activated. The page is reset to its original appearance when the anchor is de-activated.
The activation method of an anchor may be either over or
click. If the activation method is over, the anchor is
activated whenever the mouse pointer is inside the anchor area;
conversely, the anchor is de-activated when the mouse leaves the
anchor area. If the activation method is click, the anchor is
activated whenever the button is pressed inside the anchor area;
conversely, the anchor is de-activated when the button is released.
\advianchor[⟨ activation⟩]{⟨ tag⟩}{⟨ text⟩}
\begin{advianchoring}[⟨ activation⟩]{⟨ tag⟩}⟨ text⟩\end{advianchoring}
This command defines the DVI rendering of {⟨ text⟩} as an Active-DVI anchor area that plays the Active-DVI record bound to ⟨ tag⟩ when the anchor is activated.The argument ⟨ activation⟩ specifies the activation method of the anchor.
If the environment syntax form is used, ⟨ text⟩ may contain fragile commands.
Images can be encapsulated into the presentation using the Caml library CamlImages provided with the distribution of Active-DVI (see section 8.4).
Images can be lighten by specifying an alpha value (a floating point number between 0 and 1) that measures the mixing between the background and the image.
Images can also be blended, meaning that you can choose the algorithm that superimposes the image to the background. Blending modes are reminiscent of the Ghostscript blending options: the blend mode must be one of the following: normal, multiply, screen, overlay, dodge, burn, darken, lighten, difference, exclusion, (none means unset).
{\setblend{burn}
{\setalpha{0.5}
{\includegraphics[width = 0.7\textwidth]{bar.eps}}}}
In Active-DVI, colors can be specified with the conventions of the LATEX package color.sty, that is, it should either be a previously defined color or a specification of the form [⟨ model⟩] {⟨ model color specification⟩}.
For example, the following specifications are all correct:
\color{blue}
\color[named]{Yellow}
\color[rgb]{0.7,0.3,0.8}
Colors can be named using the keyword ⟨ named⟩. If you use named
colors, the color names are case sensitive and should generally be
capitalized; for instance: \color[named]{White} specifies the
white color. Hence, \color[named]{Red}{some
text} writes some text in red.
The names of available colors can be found in the dvipsnam.def
file, generally at location
/usr/share/texmf/tex/latex/graphics/dvipsnam.def.
To give an idea, the names and colors available on a standard installation of LATEX are:
GreenYellow Yellow Goldenrod Dandelion Apricot
Peach Melon YellowOrange Orange BurntOrange
Bittersweet RedOrange Mahogany Maroon BrickRed
Red OrangeRed RubineRed WildStrawberry Salmon
CarnationPink Magenta VioletRed Rhodamine Mulberry
RedViolet Fuchsia Lavender Thistle Orchid DarkOrchid
Purple Plum Violet RoyalPurple BlueViolet
Periwinkle CadetBlue CornflowerBlue MidnightBlue
NavyBlue RoyalBlue Blue Cerulean Cyan ProcessBlue
SkyBlue Turquoise TealBlue Aquamarine BlueGreen
Emerald JungleGreen SeaGreen Green ForestGreen
PineGreen LimeGreen YellowGreen SpringGreen
OliveGreen RawSienna Sepia Brown Tan
Gray Black
You may also explicitly use a CMYK (Cyan, Magenta, Yellow, Black) specification. In this case the cyan, magenta, yellow and black values follow the ⟨ cmyk⟩ keyword, and are given as a list of four integers in the range 0.0 .. 1.0. For instance, \color[cmyk]{0,1,0,0} is a valid specification for magenta.
RGB (Red, Green, Blue) specifications are similar to the CMYK specifications: following the ⟨ rbg⟩ keyword, the red, green, or blue color values, are given as floating point numbers in the range 0.0 .. 1.0. Hence, \color[rgb]{1.0,0.0,0.0} is a valid specification for red.
Active-DVI provides the package xwindows-colors, an extension to the color package, that defines a large set of the X Window System color names, as found in the file rgb.txt of a typical X installation (this file is generally located on /usr/X11R6/lib/X11/rgb.txt). To know which colors are available look at the source file of the package xwindows-colors.sty in the directory tex of the distribution.
You can modify the background of your presentation in the LATEX source of the pages. Background can be defined either as a plain color, as an image, or as a gradient (or as a combination of these!).
You can specify a global option to the background settings, so that these settings are used for the remaining pages of the presentation (otherwise the presenter resets the background options at each new page).
To modify the background of your presentation, you can:
If these options are used together, they are applied in this order: first the solid background color is drawn, then the gradient function is applied, finally the image is drawn on the resulting background. \advibg[global]{⟨ decl⟩}
where ⟨ decl⟩ is a list of settings of the following from:color=⟨ color⟩ (default value is none)
Set the background color to ⟨ color⟩. If ⟨ color⟩ is none this unsets the background color. Otherwise, ⟨ color⟩ must follow the notation above to designate colors.image=⟨ file⟩ (default value is none)
Use the image found in ⟨ file⟩ as background (none means unset).fit=⟨ fit style⟩ (default value is auto)
Fit the background image according to ⟨ style⟩, which may be one of the following keywords:auto orThe auto fit style means scaling the image as desired in both directions so that it fits the entire page. Other styles only force the same scaling factor in both directions:
topleft top topright left center right bottomleft bottom bottomright
- Corner-styles means set the image in the corresponding corner and scale it to cover the entire page.
- center means set the image in the center of the page and scale it to cover the entire page.
- Segment-styles means adjust the image and the page on the segment (in which case, the image may not completely cover the page on the opposite side).
alpha=⟨ float⟩ (default value is none)
Set the alpha channel factor for the background image to ⟨ float⟩ (none means unset). An alpha factor of 0 means that the image is not visible at all; conversely, an alpha factor of 1 means that the image covers the background.blend=⟨ blend mode⟩ (default value is none)
Set the blend mode to ⟨ blend mode⟩, which are reminiscent of Ghostscript blending options. The blend mode should be one of the following: normal, multiply, screen, overlay, dodge, burn, darken, lighten, difference, exclusion, (none means unset).gradient=⟨ function⟩ (default value is none)
Set the gradient function to ⟨ function⟩, one of the predefined functions that convert the plain background color into a color gradient from the chosen colorcolorstartto the colorcolorstop(which is white by default). Available gradients are:
hgradienthorizontal gradient (the gradient is a serie of vertical lines),vgradientvertical gradient (the gradient is a serie of horizontal lines),d1gradientfirst bissector gradient (the gradient is a serie of lines which are parallel to the first bissector),d2gradientsecond bissector gradient (the gradient is a serie of lines which are parallel to the second bissector),cgradientcentered gradient (the gradient is a serie of concentric squares),circgradientcircular gradient (the gradient is a serie of concentric circles).colorstart=⟨ color⟩ (default value is white)
Set the starting color of the gradient. When left unspecified defaults to white.colorstop=⟨ color⟩ (default value is background)
Set the end color of the gradient. When left unspecified defaults to the background color.xstart=⟨ int⟩ (default value is 0)
Set the abscissa of the lower left point of the area where the gradient is drawn.ystart=⟨ int⟩ (default value is 0)
Set the ordinate of the lower left point of the area where the gradient is drawn.width=⟨ float⟩ (default value is 1.0)
Set the width of the area where the gradient is drawn. The width is a number in the range[0 .. 1]that gives the ratio of the area width with respect to the page width (hence 0.0 means a null width and 1.0 means the entire page width).height=⟨ float⟩ (default value is 1.0)
Set the height of the area where the gradient is drawn. The width is a number in the range[0 .. 1]that gives the ratio of the area height with respect to the page height (hence 0.0 means a null height and 1.0 means the entire page height).xcenter=⟨ float⟩ (default value is 0.5)
For a centered or circular gradient, set the abscissa of the center point of the gradient into the gradient area. xcenter is a ratio of the gradient area’s width. It defaults to 0.5, meaning the middle of the gradient area width.ycenter=⟨ float⟩ (default value is 0.5)
For a centered or circular gradient, set the ordinate of the center point of the gradient into the gradient area. ycenter is a ratio of the gradient area’s height. It defaults to 0.5, meaning the middle of the gradient area height.none
Unset all background parameters. This key must appear on its own, no arguments or keys are allowed.The optional parameter global indicates that the definition is global and will affect the following pages, as well as the current page.
By default, the background settings only affect the current page.
\advitransition[global]{⟨ decl⟩}
where ⟨ decl⟩ is a list of settings of the following from:none or slide or block or wipe
Set the transition mode to the corresponding key. One of this key is mandatory (if several are provided the last one is selected).from=⟨ direction⟩
Make the transition come from ⟨ direction⟩. Directions should be one of the following:
topleft top topright left center right bottomleft bottom bottomright The default direction, to be used when no local or global direction has been specified, is determined dynamically:
rightwhen coming from previous page,leftwhen coming from next page, andtopotherwise.steps=⟨ n⟩
Make the transition in ⟨ n⟩ steps.As for \advibg, the optional parameter global indicates that the definition is global and will affect the following pages, as well as the current page.
By default, the transition definitions affect the current page only.
\advitransbox{⟨ key=val list⟩}{⟨ hbox material⟩}
where ⟨ key=val list⟩ is as above and {⟨ hbox material⟩} is whatever can follow an \hbox command. In particular, the material may contain verbatim commands, since as for the \hbox it is parsed incrementally.
The optional parameter global indicates that the definition is global and will affect the following pages, as well as the current page.
By default, the transition affects the current page only.
To animate your show, Active-DVI can launch arbitrary applications you need.
The LATEX command to launch an application during the presentation is \adviembed[⟨ key=value list⟩]{⟨ command⟩}
where ⟨ key=value list⟩ is a list of bindings of the following kind:name=⟨ name⟩
Allows to refer to the embedded application as ⟨ name⟩. Anonymous applications have actually the default name anonymous.ephemeral=⟨ name⟩
This is the default case: the application is specific to a given page. An ephemeral application is automatically launched whenever the page is displayed, and automatically killed when the page is turned.persistent=⟨ name⟩
A persistent application is launched only once and keeps running in the background; however, Active-DVI automatically hides and shows the window where the application runs, so that the application is visible only on the page where it has been launched.sticky=⟨ name⟩
A stiky application is launched only once, keeps running, and remains visible when turning pages. It is also resized and moved as necessary to fit the page size.raw=⟨ name⟩
A raw application is launched each time its embedding command is encountered. A raw application is not managed automatically by Active-DVI, except for the initial launching and the final clean-up that occurs when Active-DVI exits; hence, you can completely monitor the raw applications graphical behavior, using the advikillembed command and the window mapping facilities for raw applications described below.width=⟨ dim⟩
height=⟨ dim⟩The application takes ⟨ dim⟩ width (respectively height) space in LATEX. Both values default to 0pt.These dimensions are also substituted for all occurrences of
@gin the command string.
To monitor embedded applications, Active-DVI provides the advikillembed primitive to send a signal to any named embedded application. For raw applications, there are additional functions to map or un-map the window allocated to a named raw application. Mapping or un-mapping windows of non-raw applications is unspecified, since it may interfere in a non trivial way with Active-DVI’s automatic treatment of those applications.
Kill the embedded application named ⟨ name⟩. An optional signal value or symbolic name can be given to send to the designed process: for instance, \advikillembed[SIGUSR1]{clock} will send the SIGUSR1 signal to the embedded application named clock.Signal value defaults to -9.
Map the window of the (raw) embedded application named ⟨ name⟩.
Un-map the window of the (raw) embedded application named ⟨ name⟩.
The primitives advikillallembed, advimapallembed, and adviunmapallembed behave the same as their non-all counterparts, except that they operate on all the applications that have been launched with the given name. \advikillallembed{⟨ name⟩}
Similar to advikillembed but kill all the embedded applications named ⟨ name⟩.
Map the windows of all the (raw) embedded applications named ⟨ name⟩.
Un-map the windows of all the (raw) embedded applications named ⟨ name⟩.
Active anchors are annotated pieces of text that get associated
activation records. To define an active anchor, the command is
\advianchor[⟨ decl⟩]{⟨ tag⟩}{⟨ text⟩}
\begin{advianchor}[⟨ decl⟩]{⟨ tag⟩} ⟨ text⟩\end{advianchor}
The text is first displayed as usual, then the anchor is drawn according to the style given by ⟨ decl⟩, and made active. Its activation, which depends on the mode given by ⟨ decl⟩, will play the record named ⟨ tag⟩.The declarations ⟨ decl⟩ are of the following form:
- over, click, or stick
The mode stick plays the tag ⟨ tag⟩ on click. The mode click is similar, except that it restores the previous state when leaving the anchor area. The mode over is as click but display the ⟨ tag⟩ when the mouse is over the anchor instead of waiting for a click.
- box, invisible, or underline
this defines the style in which the anchor should be drawn. The default style is box.
In the environment form, ⟨ text⟩ may contain fragile commands.
\adviemphasize[⟨ color⟩]{⟨ text⟩}
This makes an invisible anchor around ⟨ text⟩, which when activated will redraw text in a box colored with ⟨ color⟩, which defaults to yellow.
Active-DVI can deal with most of PStricks by calling ghostscript on included Postscripts. Basic change of coordinates are implemented, but this feature remain fragile, as Active-DVI must in turn call ghostscript to get the new coordinates. Also, rotations will definitely not work for text, which is rendered by Active-DVI and cannot be rotated.
The overlay class implements overlays with PStricks. By contrast, Active-DVI implements overlays directly, using records and plays. This is more efficient, and of course more natural. (In fact, Active-DVI chooses the cumulative semantics of overlays, displaying all layers below the current overlay.)
The xprosper style, derived from the prosper class, uses
the overlay class and works with Active-DVI in exactly the same
way (relaxing the \overlay@loop macro inhibits all layers, but the first
page).
Active-DVI can deal with main PStricks. In particular, the following work
\psframe, \ovalnode, ….\Aput, Bput, etc.Other PStricks may or may not work.
Active-DVI provides this specialized LATEX package to facilitate writing presentation slides in the spirit of SliTeX. See examples in the directory examples/slitex.
This package allows superposition of horizontal material, creating the smallest horizontal box that fits all of the superpositions.
\usepackage{superpose}
The package defines a single environment: \begin{superpose}[⟨ alignment⟩]⟨ list⟩\end{superpose}
The ⟨ alignment⟩ can be one the letters c (default value), l, or r.Items of the ⟨ list⟩ are separated by
\\as in tabular environments. Each item should be a horizontal material.
This package draws bubbles over some text.
\usepackage{bubble}
\usepackage[ps]{bubble}
By default bubbles are produced using the epic and eepic packages, for portability. However, for better rendering and easier parameterization, bubbles can also be drawn using the pst-node package of the PStricks collection. This is what the ps option is designed for.
The package defines a single command: \bubble[⟨ key=value list⟩]{⟨ anchor⟩}[⟨ ps options⟩](⟨ pos⟩){⟨ text⟩}
The ⟨ key=value list⟩ is a list of bindings of the following kind:bg=⟨ color⟩ (default value is yellow)
The background color for annotations.unit=⟨ dim⟩ (default value is yellow)
Set the package unit to ⟨ dim⟩.col=⟨ colspec⟩ (default value is c)
Where ⟨ colspec⟩ is a column specification for the tabular environment. Moreover, the following abbreviations are recognized:
key expands to c col=c l col=l r col=r p=⟨ w⟩ col=p{⟨ w⟩}
key expands to C col={>{$}c<{$}} L col={>{$}l<{$}} R col={>{$}r<{$}} P⟨ w⟩ col={>{$}p{⟨ w⟩}<{$}} ⟨ pos⟩ is the optional relative position of the annotation, it defaults to 1,1, and is counted in the package units.
⟨ ps options⟩ are passed to the command \psset) in ps mode and ignored otherwise.
Parameters (color and tabular columns specifications) can also be set globally using the command: \bubbleset{⟨ key=value list⟩}
This package uses active anchors and the bubbles package to provide annotations by raising a bubble when the cursor is over the anchor.
The package defines a single command \adviannot[⟨ key=value list⟩]{⟨ anchor⟩}[⟨ ps options⟩](⟨ pos⟩){⟨ text⟩}
whose options are identical to those of the \bubble macro; however the bubble appears within an active anchor.
This 3-lines long package loads the graphicx.sty package and provides declarations so that JPEG, EPS, TIF, TIFF source images can be embedded: Active-DVI will preview these images directly while other drivers will translate them on demand.
Postscript fonts are not natively handled by Active-DVI. You must use the command dvicopy to expand those virtual fonts to base fonts before visualization with Active-DVI. (For instance, dvicopy talk.dvi talk.expanded; advi talk.expanded very often does the trick.)
PS relies on ghostscript to display Postscript in-lined specials. However, some earlier releases of ghostscript implements the Postscript flushpage command as a XFlush call which does not force the evaluation of commands, and thus makes the synchronization between ghostscript and Active-DVI drawings uncontrollable. In this case, the interleaving of in-lined postscript and other material may be inconsistent.
Fortunately, recent versions of ghostscript (> 6.5) have fixed this problem by using XSync(false) instead. If you use those versions of ghostscript, in-lined specials should be correctly rendered.
Unfortunately, some releases of version 6.5x also carry a small but fatal bug for Active-DVI, that will hopefully be fixed in future releases. A workaround is available here http://cristal.inria.fr/~remy/ghostscript/.
So far, the implementation of in-lined Postscript does not correctly handle complex change of coordinates. (See PStricks section).
Please, send bug reports to
mailto:advi@inria.fr.
See http://gallium.inria.fr/advi for up to date information.
Active-DVI recognizes the keystrokes listed below when typed in its window. Some keystrokes may optionally be preceded by a number, called ARG below, whose interpretation is keystroke dependant. If ARG is unset, its value is 1, unless specified otherwise.
Active-DVI maintains an history of previously visited pages organized as a stack. Additionnally, the history contains marked pages which are stronger than unmarked pages.
| ? | info | – | This quick info and key bindings help. |
| q | quit | – | End of show. |
| space | continue | – | Move forward (ARG pauses forward if any, or do as return otherwise). |
| ^X-^C | quit | – | End of show. |
| n | next | – | Move ARG physical pages forward, leaving the history unchanged. |
| p | previous | – | Move ARG physical pages backward, leaving the history unchanged. |
| , | begin | – | Move to the first page. |
| . | end | – | Move to the last page. |
| g | go | – | If ARG is unset move to the last page. If ARG is the current page do nothing. Otherwise, push the current page on the history as marked, and move to physical page ARG . |
| N | next pause | – | Move ARG pauses forward (equivalent to continue). |
| P | previous pause | – | Move ARG pauses backward. |
| ^X-^F | set fullscreen | – | Adjust the size of the page to fit the entire screen. |
| ^F | toggle fullscreen | – | Adjust the size of the page to fit the entire screen or reset the page to the default size (this is a toggle). |
| < | smaller | – | Scale down the resolution by scalestep (default (((2)1/2)1/2)1/2). |
| > | bigger | – | Scale up the resolution by scalestep (default (((2)1/2)1/2)1/2). |
| # | fullpage | – | Remove margins around the page and change the resolution accordingly. |
| c | center | – | Center the page in the window, and resets the default resolution. |
| h | page left | – | Moves one screen width toward the left of the page. Does nothing if the left part of the page is already displayed |
| l | page right | – | Moves one screen width toward the right of the page. Does nothing if the right part of the page is already displayed |
| j | page down | – | Moves one screen height toward the bottom of the page. Jumps to the top of next page, if there is one, and if the bottom of the page is already displayed. |
| k | page up | – | Moves one screen height toward the top of the page. Jumps to the bottom previous page, if there is one, and if the top of the page is already displayed. |
| move page | – | A black line draws the page borders; moving the mouse then moves the page in the window. | |||
| ^C |
| – | Toggles center-on-cursor flag, which when sets moves the screen automatically so that the cursor appears on the screen. |
| w | switch | – | Switch view between master and client (if any). |
| W | sync | – | Goto page of client view corresponding to page of master view. |
| ^W | autoswitch | – | Toggle autoswitch flag. |
| r | redraw | – | Redraw the current page to the current pause. |
| R | reload | – | Reload the file and redraw the current page. |
| ^L | redisplay | – | Redisplay the current page to the first pause of the page. |
| a | active/passive | – | toggle advi effects (so that reloading is silent). |
| return | forward | – | Push the current page on the history stack, and move forward n physical pages. |
| tab | mark and next | – | Push the current page on the history as marked, and move forward n physical pages. |
| backspace | back | – | Move ARG pages backward according to the history. The history stack is poped, accordingly. |
| escape | find mark | – | Move ARG marked pages backward according to the history. Do nothing if the history does no contain any marked page. |
| T | Thumbnails | – | Process thumbnails. |
| t | toc | – | Display thumbnails if processed, or floating table of contents if available, or else do nothing. |
| s | write | – | Give a pencil to scratch, typing characters on the page. |
| S | draw | – | Give a spray can to scratch, drawing on the page. |
| ? | info | – | While in scratch mode, press ? for more info. |
| ^X-l | toggle laser | – | Toggle the laser beam to point on the page. |
| ^G | laser off | – | When laser is on turn it off. |
| ^X-^S | save page | – | Save the current page as an image file. |
| f | load fonts | – | Load all the fonts used in the document. By default, fonts are loaded only when needed. |
| F | make fonts | – | Does the same as f, and precomputes the glyphs of all the characters used in the document. This takes more time than loading the fonts, but the pages are drawn faster. |
| C | clear | – | Erase the image cache. |
Press s to enter scratch writing; the cursor is modified and you must click somewhere on the page to start writing text there. Before clicking, you can
Active-DVI recognizes the following keystrokes when scratch writing on the page.
| ^G | quit | – | End of scratch writing. |
| Esc | settings | – | Enter the scratch writing settings mode. |
In the scratch writing settings mode, the cursor is modified and you can set some charateristics of the scratch writing facility. When in doubt, press
When in the scratch writing settings mode, the following keys have the following respective meanings:
| > | greater | – | Increments the scratch font size. |
| < | smaller | – | Decrements the scratch font size. |
| b | blue | – | Set the color of the font to blue. |
| c | cyan | – | Set the color of the font to cyan. |
| g | green | – | Set the color of the font to green. |
| k | black | – | Set the color of the font to black. |
| m | magenta | – | Set the color of the font to magenta. |
| r | red | – | Set the color of the font to red. |
| w | white | – | Set the color of the font to white. |
| y | yellow | – | Set the color of the font to yellow. |
| B | more blue | – | Increment the blue component of the color. |
| G | more green | – | Increment the green component of the current color. |
| R | more red | – | Increment the red component of the current color. |
| + | positive increment | – | Set the color increment to positive. |
| − | negative increment | – | Set the color increment to negative. |
| ? | help | – | Give the list of settings available. |
| Esc | quit | – | Quit the sratch writing settings mode. |
Just press Esc to enter the scratch writing settings mode, then > or < to increment or decrement the font size; then press Esc again, to leave the scratch writing settings mode and continue to write on the page with the new font size.
Press S to enter scratch drawing; the cursor is modified and you must click somewhere on the page to start drawing there. Before clicking, you can
Active-DVI recognizes the following keystrokes when scratch drawing on the page.
| ^G | quit | – | End of scratch drawing. |
| Esc | settings | – | Enter the scratch drawing settings mode. |
In the scratch drawing settings mode, the cursor is modified and you can set some charateristics of the scratch drawing facility.
When in the scratch drawing settings mode, the following keys have the following respective meanings:
| b | blue | – | Set the color of the font to blue. |
| c | cyan | – | Set the color of the font to cyan. |
| g | green | – | Set the color of the font to green. |
| k | black | – | Set the color of the font to black. |
| m | magenta | – | Set the color of the font to magenta. |
| r | red | – | Set the color of the font to red. |
| w | white | – | Set the color of the font to white. |
| y | yellow | – | Set the color of the font to yellow. |
| B | more blue | – | Increment the blue component of the current color. |
| G | more green | – | Increment the green component of the current color. |
| R | more red | – | Increment the red component of the current color. |
| + | positive increment | – | Set the color increment to positive. |
| − | negative increment | – | Set the color increment to negative. |
| > | increment | – | Increment by one the size of the line. |
| < | decrement | – | Decrement by one the size of the line. |
In the setting mode, pressing one of the following keys enter the (still experimental) figure drawing mode:
| V | vertical line | – | Draw a vertical line. |
| H | horizontal line | – | Draw a horizontal line. |
| S | segment | – | Draw a segment. |
| C | circle | – | Draw a circle. |
| p | point | – | Draw a point. |
| P | polygone | – | Draw a polygone. |
| e | endpoly | – | Close the polygone that is beeing drawn. |
| F | free hand | – | Draw a line following the pointer. |
| ’ ’ | cancel | – | Cancel the figure setting. |
|