conception of a DSKY

What is a DSKY?

The DSKY is the "display/keyboard" (DSKY) used by the Apollo Guidance Computer (AGC).  The DSKY provided only a means to input keyboard data to the AGC, or to display visual information at command of the AGC, and therefore had little or no functionality of its own, when considered as a stand-alone device. 

The  DSKY may seem somewhat familiar from, for example, the movie Apollo 13. In the HBO mini-series, From the Earth to the Moon, the DSKY appears so prominently that it almost becomes another character (dramatically speaking). For those of us who are familiar with it from these sources, it is easy to think of the DSKY itself as being the computer.  However, the DSKY really was quite independent from the AGC.  From the computer's standpoint, the DSKY was just another few i/o channels, more or less, as in the block diagram below.

Click to

Guidance&Navigation block architectureor

The Lunar Module had a single DSKY, at the bottom center of the control panel.  The Command Module had two DSKY units: one in the main control panel near the commander; the other at the navigator's station.
              control panel.Real
              DSKY.Guidance, navigation, and control system.CM control panel.

One thing you may wonder about if you looked at the pictures above in a particularly scrutinizing fashion is this:  Why, in the half-dozen or so pictures of the DSKY above, are there four or five different collections of words displayed on the upper-left panel of the DSKY?  Or perhaps you didn't even notice, since people are so used to graphical displays that simply show whatever the software requires ... so what difference does it make what they display?  Except that the DSKY didn't have a graphical display.  Its design long preceded the notion of a useful graphical display.  The words shown on the DSKY's front panel were actually engraved on there.  They couldn't just change at will! 

Perhaps the point still isn't clear even after that explanation, so look at the image to the left, which I've extracted DSKY annunciator-lamp arrangements from the various images above.  The two leftmost arrangements of annunciator lamps were actually never flown at all ... although if you squint very closely at the October 1966 photo of a Command Module control panel above, you'll see that the leftmost arrangement of annunciator lamps does seem to have been used in its DSKY.  With that said, the rightmost arrangement was invariably used in the flown Command Modules, and was used in the Lunar Module up through Apollo 10, while beginning with Apollo 11 a different arrangement (not depicted) began to be used in the Lunar Module.  By the way, I once asked Eldon Hall what the leftmost arrangement was actually used for, since it was never flown, but nevertheless still appears quite often in print articles?  He simply said that the faces were modular, and thus interchangeable, leaving me the impression he thought I was an idiot.  (But I think the question is still valid.  )

In fact, what this illustrates is that "the DSKY" is not just one thing, but rather a sequence of stages of a design process.  Occasionally, people take me to task for ignoring such details, not to mention other specifics like the exact dimensions of the DSKY, the exact colors of the faceplate or the indicator lamps, and so forth, simply because such details were not available when I began the project of simulating the AGC and DSKY in software.  Fair enough, though I'm going to basically continue ignoring such details below.  Before I revert to doing that, though, I'd like to point out that we now have quite a collection of the original engineering drawings that define the DSKY design, so many of these questions can be answered by the engineering drawings rather than requiring a trip to the Smithsonian with calipers and whatever device is used to measure color.  In fact, here are links to the engineering drawings for all Block II DSKYs flown on missions:

Which still leaves my question to Eldon Hall unanswered, about where the leftmost arrangement of indicator lamps was used?  I should leave that as an exercise for the reader, but I'll take pity and just give you the answer:  If you are persistent in working through the engineering drawings, you find this arrangement in the DSKY for the very earliest Block II G&N (Guidance & Navigation) system, which was designated G&N system #201, and specifically in drawing 1006316, entitled "INDICATOR, ALARM, SPECIFICATION CONTROL DRAWING".  This drawing (and others like it for other variations of the G&N system) are also full of other details of interest to those concerned about authenticity of physical appearances, such as that the indicator lamps shall be "aviation white" and "aviation yellow" per mil-spec MIL-C-25050.  One interesting factoid asserted by the drawing is that the lettering shall be 0.156" high black Gorton Condensed.  Other versions of the DSKY instead call out Gorton Normal.  You may suppose from this that "Gorton" was some generally recognized font.  It was not.  Gorton was in fact the name of a manufacturer of engraving machines.  However, Gene Dorr (who is actually the one who researched these matters) has gone ahead and created an actual Gorton font, which could be of use to anyone trying to create their own, far more authentic-looking version of a DSKY than I have! 

As for G&N system #201, I suppose that it was simply such an early version that the specific indicator lamps needed hadn't been set in stone yet, but that once the media had an attractive graphic it simply ended up being reproduced in print articles.  As far as the middle arrangement shown above, I find no record of it in any of the engineering drawings I've seen.

What is yaDSKY?

yaDSKY and DSKY LITE, side-by-sideyaDSKY is a computer program which emulates the DSKY.  yaDSKY has now been superceded by yaDSKY2, though both programs are still available, and with only minor differences are interchangeable.  So in speaking of "yaDSKY" I am really speaking of either program unless I explicitly state otherwise.

The DSKY had little or no functionality of its own, when considered as a stand-alone device.  The same is true of the yaDSKY emulation of the DSKY.  It requires the yaAGC program (which simulates the AGC) to be useful.    Just as the true AGC and DSKY communicated between themselves by means of wiring, the virtual yaAGC and yaDSKY communicate between themselves using communication channels (sockets) that act like virtual wires.   In other words, yaDSKY merely communicates keypresses to yaAGC, and receives communications from the yaAGC about which lights to activate.

If you don't like my yaDSKY, or if you want to integrate yaAGC into a larger simulation project, or if you want something that works on a platform I don't support, you can replace yaDSKY with a program more to your liking. The method for doing so is described in detail on the developer-info page.  In fact, Stephan Hotto has done just that (among other things) with the LM_Simulator software he has contributed to this project.  At the right is a very cute screenshot (click to enlarge) of yaDSKY and Stephan's "DSKY LITE" running side-by-side and communicating with the same instance of yaAGC.

Running yaDSKY

The yaDSKY2 program is invoked as:


while yaDSKY is invoked as

yadsky [OPTIONS]

The available options are as follows:

Display a list of options, something like the information shown here.

yaDSKY and yaAGC are a client/server pair, and can be running on the same computer or on different computers. In order to connect to the yaAGC server, yaDSKY has to know the IP-address of the machine on which yaAGC is running. By default this is "localhost"—i.e., the two are running on the same computer. However, a different IP address may be specified using this command-line switch. In Linux, this may be either the machine's name, or else a numerical address in dotted format (n.n.n.n). MS-Windows—at least Windows 98—is somewhat crippled in this respect, and will only accept a host name rather than a numerical IP address.

Similarly (see above), in addition to an IP address, a port number is needed. This must be in the range of port numbers yaAGC is scanning (which by default is 19697-19706).  If no port is specified, yaDSKY will use 19697 (since the Apollo 11 mission was in July 1969). A simulation of the Command Module would require two DSKY units, one for the contral panel and one for the navigator's station; so two copies of yaDSKY would be run, and would have to be assigned different ports (such as 19697 and 19698).  The developer page contains suggested port assignments, including the suggestion that Colossus should use the range 19697-19706 while Luminary should use the range 19797-19806.

This switch configures the DSKY for different Apollo missions or for CM vs. LM operation.  Among the things that differ in different configurations are the legends that appear on the 14 indicator lights on the upper-left side of the DSKY panel, and the labeling on the PRO key.  The DSKY is reconfigured through the use of configuration files, of which the ones that exist as of this writing are CM0.ini, CM.ini, LM.ini, and LM1.ini.  (A description of the differences appears later.)  If no "--cfg" switch is present, default values are used which are the same as those in the distribution version of LM.ini, but the LM.ini file itself is not used unless specified by the presence of "--cfg=LM.ini".  Refer to the developer page for information on modifying the legend sets or adding new sets.   If a complete pathname for ConfigFilename is not specified, then yaDSKY will look first in the current directory, and then (for software versions 05/06/2004 or later) in the installation directory (which is /usr/local/bin by default, but which may be changed at build-time).

(Only versions 20040811 and after.)  yaDSKY's graphical interface is simply too big for PC's with lower graphical resolution (smaller than 1024×768).  If the --half-size switch is used, half-size graphics are used, and so the interface should be usable even down to 640×480 resolutions.  The smaller interface doesn't really look very good, because it is simply blindly scaled down from the larger interface, rather than being optimized, but at least it's usable at 800×600 and 640×480 resolutions.  If both the --half-size and --cfg switches are used, the --half-size switch must be to the left of the --cfg switch on the command line.

(Versions 20040821 and after.)  Adds a delay at start-up, so that yaDSKY does not immediately begin attempting to communicate with yaAGC.  The current defaults are 0 ms. in Linux and 500 ms. in Win32.  This "feature" has been added as a temporary work-around for problem report #23, and probably has no other sensible purpose.  Even on Win32 it isn't usually needed, but it's here for the 10% (or whatever) of the time it's needed.

(Versions 20050515 and after.)  This is present only for debugging the AGC's "unprogrammed" counter-increments, and has no useful purpose otherwise.  In this mode, the keys of the simulated DSKY—or at least, the VERB, NOUN, 0-7, and PRO keys—do not have their normal interpretations and cannot be used to affect the AGC CPU in any normal way.  Instead, the combination NOUN-digit-digit is used to specify an (octal) CPU counter-register number, and the combination VERB-digit is used to select a counter-increment type as defined on the developer page.  The PRO key is used to send the selected counter-increment command to the AGC, to increment/decrement the counter-register.  Note that there are no visual displays associated with this, since the AGC is still commanding the DSKY visual display in its usual way, so you just have to mentally keep track of what counter-register and which unprogrammed-increment command you've chosen.  This mode is terminated only by restarting yaDSKY.  Only counter registers in the range 32-60 (octal, default 32) are accepted, and only increment-types in the range 0-6 (default 1) are accepted.  Any time the PRO key is pressed with settings outside this range, no action occurs.  The only useful way to use this is to make sure that yaAGC has been run in debug mode, and to use the yaAGC debugger's "dump" command to check the results of the various increment/decrement commands.  For example, if the default settings aren't changed, the command "dump 32" will initially show 00000, but will show "00001" after the PRO key has been hit once, "00002" after the PRO key has been hit twice, and so on.

(Versions 20050626 and after.)  In this mode, keystrokes are communicated to yaAGC via the digital uplink rather than via the normal CPU input channel 015, but this is relatively transparent to the user, other than the fact that the UPLINK ACTY indicator lights up.  Normally this facility would have been used by the ground-station for uplinking data.  However, since the uplink consists of encoded DSKY keycodes, "--test-uplink" provides a convenient method of testing this facility without having to write a whole new program.  I've noted that when using this mode, the keystrokes seem to work properly, but updating the DSKY display is very hit-and-miss.  I assume this is a normal property of the uplink—rather than a bug in yaAGC—but I don't really know. 

(yaDSKY versions 20050628 and after.  Not available in yaDSKY2, but equivalent functionality is provided instead by the separate yaTelemetry program.)  In this mode, yaDSKY provides the additional function of displaying telemetry-downlink lists as they are emitted from the AGC, in a manner similar to the yaTelemetry program.  To use this mode effectively, yaDSKY should be run from a terminal window which is at least 80 text-columns wide, and at least 36 text-rows high; otherwise, a complete downlink list will not fit on-screen and portions will scroll off the top of the screen.  Furthermore, an ANSI terminal will provide the most effective (and attractive) display, since some ANSI escape-sequences are used for cursor positioning.  In Linux this should be no problem.  In Windows, it is generally necessary to install ANSI.SYS in order for the command-line interface to accept ANSI commands.  I must admit, though, that in experimenting with Windows XP I have been unable to get ANSI.SYS to install.  (If somebody wants to send me the magic formula for this, I'd appreciate it.)

(Final yaDSKY version only.  Not available in yaDSKY2.)  Alters the locations of the files containing graphics files used by the yaDSKY program to locations more congenial to use in VirtualAGC GUI based installations as opposed to installation in system directories (the default).  The new locations used are the ./pixmaps/yaDSKY/ directory.

yaDSKY and yaAGC attempt to be insensitive to each other's absence, but I'd recommend running yaAGC first and yaDSKY second, particularly in Win32.

Features for Presentations, Interactive Demos, etc.

Other than the techniques discussed in this section, it may be possible to use the digital uplink facility to automate various activities.

Canned, Pre-recorded Mission Segments

(This section is applicable to software versions 2017-12-11 or later.)

It is possible to record (or otherwise create) a script of AGC i/o-channel operations controlling the DSKY, and to conveniently play back such canned scripts later.  At present, there are two ways of obtaining such canned scripts:

In playback, yaDSKY2 basically disregards any new output-channel commands originating from the AGC in real time, working only from the script it is playing back.  When the script has completed or perhaps has been prematurely terminated by the user, it restores the DSKY's numerical displays and indicator lamps to the configuration they had prior to starting the playback, and resumes listening to the AGC for new commands.  Usually the AGC commands that were ignored during script-playback will have been buffered and will all be processed instantly, so that the DSKY will quickly assume the appearance it would have had if it had been responding to the AGC's commands all along.

At the moment I'm writing this, we have the following pre-recorded scripts, each taking 15-16 minutes to run:

To select a script to play back, click the yaDSKY2's PROG indicator lamp with the mouse.  Of course, in a real DSKY, the PROG indicator and all other indicators are simply lamps, and not buttons, and it's only the simulated DSKY in which some indicators have been subverted to have button-like properties.  At any rate, doing so opens up a file-dialog that allows you to select the canned script to be played back.  Usually these will be in the directory the file dialog shows by default, though if you have created canned scripts of your own, they will be located elsewhere and you'll have to use the file dialog to navigate to them.  The files have the extension ".canned".

Scripts will generally execute all the way to their ends, displaying a pop-up information box when complete, but you can also prematurely terminate a script by once again clicking PROG whilst the script is running.

A script is simply a text file, and therefore can in theory be created by some other method than the ones I will proceed to describe below.  The format is very simple; each line simply has three fields,

DifferentialTimeInMilliseconds AgcOutputChannelNumberInOctal AgcOutputChannelValueInOctal
The differential time (usually an integer, but can be floating-point decimal) is simply the time in milliseconds since the preceding line of the file.  (NASSP i/o-channel logs have the same format, except that the initial field is the absolute simulation time, in seconds.)  DSKYs are always controlled by AGC "output channels" 010, 011, 013, and 0163 as described on our developer page; for informational purposes, input channels that record keystrokes (outputs from the DSKY) are generally included as well.

To record a file from yaDSKY2 (or for later playback, you'd do this:

Finally, it is sometimes useful to analyze these i/o-channel script in order to understand or document what they're doing.  There's a program called that produces a relatively human-friendly report from a playback script.

"Special Effects"

If you should find yourself in the position to give a presentation on the AGC, the lunar landings, Apollo, etc., using Virtual AGC, versions of the yaDSKY2 program 20090613 and later have a "special effects" facility build into it which might be useful to give your presentation a little more sizzle.  Conceivably, this facility could also be used in setting up an interactive AGC demo, AGC self-study materials, etc.  What the facility allows you to do is to map selected sequences of DSKY keystrokes to commands which will be executed on the computer running the simulation.  The only real restrictions are that the key-sequences used must begin with the VERB key and cannot contain a second VERB key in the sequence; also, at most 100 keypad-to-command mappings can be defined.

This keystroke-to-command mapping is independent of the processing done by the AGC.  In other words, if a certain keystroke pattern is configured to execute a certain command on the simulating computer, it will still cause the AGC to perform whatever action the AGC would normally perform given those keystrokes.

In order to configure yaDSKY2 to do this, you need to create a file called "DSKY2.matches" in the Resources/ subfolder of the installation folder.  (We discuss this elsewhere, but for default installations that folder is "~/VirtualAGC/Resources/" in Linux, "c:\program files\virtual agc\resources" in Windows, and "~/Desktop/" in Mac OS X.)  This ASCII file is simply a sequence of lines which begin with the keystroke pattern being matched, followed by a space, followed by the command which is supposed to be executed.  The command can have spaces or pathnames in it.  The patterns are created from the characters VN+-0123456789CPKER (capitalization is important!), which correspond to the DSKY keys in an obvious way (see the following section).  The pattern must begin with V and contain no more than one V, since the match buffer is cleared whenever the VERB key is pressed.  You can also use the special pattern "startup", in which the associated command is executed when yaDSKY2 starts up rather than having to be activated by any particular sequence of keystrokes.

Basically, anything which can be controlled from the computer's command-line can be controlled by a DSKY key sequence in this way.   What these things are depends on the operating system, the peripherals installed on your computer, and the ease of controlling those peripherals from command-lines (as opposed to GUI controls)—not to mention your own creativity and computer skills—so I won't bother with listing any specific commands you might use.  But in a general way, you might:
For example, suppose that your PC was configured so that it could control an X10 automation system which could (among other things) turn off the room lights by executing the command-line command "LightsOff.bat", and that to begin your presentation you wanted to turn off the room lights and start a short introductory video; later in the presentation you want to manually demonstrate some AGC operations, and you want to trigger some special action when performing a "goto pooh" operation ("V37E00E").   As mentioned, the V37E00E command is still processed properly by the AGC and still causes program P00 to start.  For this example, your DSKY2.matches file might look like this:

startup LightsOff.bat
startup PlayIntroVideo.bat
V37E00E GotoPoohAction.bat

This example assumes that you are running Windows (since that's the only place that a ".bat" file is used; on other supported platforms, the equivalent concept would be a "shell script").  You provide the batch files yourself, and yaDSKY2 has no understanding at all of what they represent, other than that it is supposed to execute them.  This example needs only a single keystroke mapping, but if there were additional keystroke mappings, you would simply add more lines to the file.  Also, while we illustrate this with batch files (shell scripts), the commands could just as easily be executable programs with command-line parameters as well.

It's very important to understand that yaDSKY2 has no means of locating the commands referenced by DSKY2.matches other than the the local rules (such as the PATH environment variable) used by your system.  Furthermore, the setting of PATH isn't necessarily the same when you are running VirtualAGC by clicking its desktop icon than it is when you are testing out batch files or shell scripts from a command line.  Therefore, it may be best to specify the full pathnames of the commands being used rather than just the simple filenames as shown in the example above.  Personally, I place these command files in the same folder as DSKY2.matches itself, and specify the names as "./filename" or ".\filename".

In the future, this facility could be extended in various ways, such as triggering commands based on the DSKY indicator lamps becoming active, numerical patterns appearing in the DSKY display registers, adding commands for initiating digital uplinks, etc.  However, it is unlikely that any of these capabilities will be implemented unless there are explicit requests for them, so contact me directly if you have a need for some or all of them.

The Different Parts of the DSKY

It's important to understand that the various keys and displays of the DSKY have no intrinsic interpretations of their own.  The interpretations of the keys and displays are entirely dictated by the software running on the AGC; with different AGC software, different interpretations might apply.  Fortunately, though, the software was fairly consistent in this regard, and the typical meanings of the keys and displays are summarized below.

The Keys

0-9, +, -
Self-explanatory, I think.
"Verbs" and "nouns" were used by the astronaut to supply commands to the AGC. First a verb would be entered, then a noun. (For example, verb 37 would indicate that the astronaut desired the AGC to run a particular sub-program. The noun would be the requested sub-program number.) In general, the sequence of steps for keying in a noun/verb pair was: VERB, digit, digit, NOUN, digit, digit, ENTR.
Clears the current data display.  Pressing CLR twice clears two data displays.
This key was pressed by the astronaut to "release" the DSKY. In other words, to tell the AGC that he was done with what he was doing and that the AGC could begin displaying other data on the DSKY, if it desired to do so.
Toggles "standby" mode on or off.  In other words, if in normal operating mode proceeds to standby mode, but if in standby mode proceeds to normal operating mode.
Resets (turns off) the indicator lamps.

In yaDSKY, you use the mouse to click the various buttons on the keypad.  For yaDSKY versions 2005-09-20 and later—but not in yaDSKY2—thanks to Christian Bucher you can also use the PC's keys

0 1 2 3 4 5 6 7 8 9 + - v n e

(DSKY keys CLR, KEY REL, PRO, and RSET have no accelerator keys on the PC's keyboard.)  For yaDSKY2, alas! these hotkeys presently work only in Mac OS X, and don't work either in Linux or Windows.

The Displays

This lit up when the computer was busy.
The PROG lamp was normally lit. The two-digit display underneath it showed the program number that was currently running. The Luminary and Colossus programs each provided up to 100 sub-programs for various purposes. The sub-programs were numbered 00, 01, ..., 99.
The VERB lamp was normally lit, but flashed when the astronaut was supposed to key in a new verb number. The last verb keyed, or the verb currently being keyed, was shown on the two-digit display beneath the VERB lamp.
Same idea as the VERB display, but for nouns instead.
Register 1, 2, and 3
Three 5-digit displays (plus sign) showed data that was dependent on the particular sub-program being run at the time. The numbers were decimal if the plus or minus sign was lit, but were in octal format otherwise.

The Warning or Indicator Lamps

The DSKY panel had space for up to 14 warning lights in the upper-left quadrant of its panel.  The trick is determining the textual legends for the indicators, and the colors of those indicators, on a mission by mission basis.  The data for doing so comes from the engineering drawings for the DSKY, which can be summarized as follows:

However ... the drawings referenced above are misleading for Apollo 15-17, because what apparently happened after the DSKY was manufactured per the drawings is that the two indicator lamps left blank by the engineering drawings were supplemented by decals assigning them legends after all.  (Or perhaps there were just later revisions of the engineering drawings than those we've been able to acquire 50+ years after the fact, and the later revisions showed the decals.)  The Luminary 1E Guidance System Operations Plan document describes the decals thusly:

NO DAP Light - (Placarded NO DAP - opposite VEL light)
PRIORITY Light - (Placarded PRIO DISP - opposite ALT light)

As for coloring, the ALARM INDICATOR drawings mentioned above imply the colorings:

The drawings (and descriptions of added decals) also say that the following physical arrangements of text&color apply:

CM, all Block II missions
LM, Apollo 5,9-10

LM, Apollo 11-14


LM, Apollo 15-17

× × × Lit whan data was being received from the ground
× × × × Lit when the temperature of the stable platform containing the sensors for the inertial measurement unit (IMU) was out of tolerance.
× × × × Lit when the inertial subsystem could not provide attitude reference.
× × × × Lit when the middle gimbal angle (for the IMU stable platform) was greater than 70 degrees.
× × × × Lit when the computer system was in standby.
× × × × Lit when the computer was waiting for operator input.
× × × × Lit when the computer wanted to display some information on the DSKY, but was locked out from doing so because the crew were using the DSKY. The astronaut was supposed to press the KEY REL key when this light came on, to indicate that he was done with the DSKY.
× × × × Lit while the computer was restarting.
× × × × Lit to indicate a data-entry error by the crew.
× × × × Lit when one of the optical coupling units failed.

Lit when PGNS DAP is not controlling attitude (minimum impulse mode or idling mode).

Lit when AGC attempts to display a priority display.  How this differs from KEY REL has been described to me as follows:  For KEY REL the astronaut decides when to see what the computer is waiting to display; whereas for PRIO DISP, the computer decides to display something and then the astronaut decides when to dismiss that presentation (PRO, V33E, V32E, V34E).

× ×
Lit when the altitude was out of range.

× ×
Lit when the velocity was out of range.




This page is available under the Creative Commons No Rights Reserved License
Last modified by Ronald Burkey on 2021-11-03.

Virtual AGC is hosted