TilEm2 is a TI calculator emulator. It emulates all the Z80 calculators (73, 76.fr, 81,
82, 82stats, 82stats.fr, 83, 83+, 83+ SE, 84+, 84+ SE, 85, and 86) and all known
ROM/OS versions.
TilEm2 is completely free, and designed for Linux (but available for Windows).
We put a lot of work in this software to offer to the community the best possible
product.
TilEm2 also provides a full featured debugger with disassembler, breakpoints,
memory view and more.
Some of you probably already know TilEm because a first version was released
around 2000/2001 by Julien Solignac (then maintained by Benjamin Moody since
2004).
This first version was working fine but there were some issues, skins were too small
and bad resolution and a lot of feature were missing.
Anyway, this software was pretty good (especially because the core emulation was
very good).
We decided to rewrite this emulator from scratch, keeping the philosophy of
TilEm but improving all the rest.
A new core has been developped by Benjamin Moody (aka ”floppusmaximus”), and I
(Thibault Duponchelle aka ”contra-sh”) started to work on the GTK user interface
(later he helped me for this task).
We are proud to release our work for beta testing !
TilEm2 has basically all the TilEm old features plus a lot of new things
:
Here’s the right click menu option :
TilEm2 do a lot of stuff that TilEm1 was not able to do, but there’s always some
feature not implemented (yet).
But do not forget that developpement goes on and we are planning to do it
!
You can use TilEm2 without skin (just uncheck the ”Use skin” checkbox into the
Preferences menu) but skins are more user friendly :)
We have made some officials and free to use skins (thank you to our contributors).
You can do your own skins using skinedit. If you want, you can send us the skin file,
maybe it could become ”official”.
What do you need to do your own skin?
Just take a picture of your calc using your smartphone by example.
Scale it keeping proportion to have around 900 pixels high.
Then start skinedit, create a new skin, open the picture and set the key
positions.
It takes less than 20/30 minutes I think.
See the chapter ”Create your own skins” to know how to do.
Then you can test it with TilEm2.That’s all!
Here are the current skins available by default :
Before installing TilEm2, you should know that no ROM is included in this
software.
In order to use TilEm2, you must use your own rom (use TILP to get it).
TilEm2 provides an installer msi for windows and script autoconf for Linux.
There’s not a lot of dependancies so you should really have no problem to install
it.
TilEm2 uses the following libraries :
You can find libticalcs2 on ticalc (from Romain Lievins).
Download the sources of TilEm2 on sourceforge.net.
Or eventually :
Dowload the source from the trunk like this :
Then install gtk+ (e.g. for debian : sudo apt-get install libgtk2.0-dev).
Then install libticalc2.
(http://www.ticalc.org/archives/files/fileinfo/374/37479.html)
After that, simply use the configure script and the well know Linux install
:
If you have no errors, so dependancies are checked and it’s ok.
The Makefile have been generated so type:
Then to copy the icons, configuration files and tilem2 binary type:
Usually, icons will be copied into /usr/share/tilem2/
Keybindings and configuration file will be installed into $HOME/.config/tilem2
Then you can launch TilEm2 with the command :
Or simply :
If you do not specify explicitely a rom on the command line, the first launch will ask
you which rom you want to use.
If you don’t know what’s a rom or how to get it, please read the chapter ”Get a
rom”.
TilEm2 will open a file chooser dialog.
As soon you press ”Ok”, TilEm2 will try to guess the model of this rom (and
check if it’s a correct rom).
When TilEm2 has a doubt, he will ask you for the model but will display only the
possible candidates (not all the z80 calc).
Anyway, if you launch a rom, you usually know what’s model it is because as I’ve
already said : you should have the calculator of the rom you’re trying to
emulate...
After that, TilEm2 start (but calc is off you need to press on).
The last used rom is automatically used for the next launch of TilEm2.
You can save the current state of the calculator by using ”Save Calculator”.
If you cancel the rom chooser dialog, TilEm2 automatically shutdown.
In order to emulate a calculator, you must have a copy of the calculator’s operating system. This file is referred to as a “ROM image,” because, traditionally, the OS was stored in the calculator’s Read-Only Memory. (More recent calculator models store the OS in Flash memory instead, but the name has stuck.)
The ROM code (which forms the “brain” of the calculator) is copyrighted by TI, so it is not included with TilEm. Instead, you will need to copy the ROM from a calculator you own. There are various ways to do so:
TiLP is a free software program for communicating with TI calculators of all types. Like TilEm, it runs on a variety of platforms, including Windows, Mac OS X, GNU/Linux, and FreeBSD. (If you’ve installed TilEm from source, you’ve already done most of the work of installing TiLP.) You can download TiLP from http://tilp.sourceforge.net/.
To use TiLP, you will also need a cable to connect your calculator to your computer. The TI-84 Plus can be connected using a standard mini-USB cable; for other calculators, you can either buy an official TI Graph Link cable, or (if you’re feeling adventurous) build your own.
When you launch TiLP, the main window appears:
If it does not detect your calculator automatically, select File ▹ Change Device (for older versions of TiLP, right click on the left pane and select Change Device.) Ensure that the “Cable” and “Calc” options are correct, and that your calculator is turned on and connected. Click the Dirlist button to list the variables on your calculator, and test whether the cable is working.
Dumping the ROM requires you to run an assembly program on your calculator. Although these programs have been well tested, there is always the possibility of something going wrong, so you may want to take this opportunity to back up any important files from your calculator.
If you have a TI-73, TI-82 or TI-85, you will first need to install an assembly shell. These shells are not included with TilEm or TiLP, but can be downloaded from the Web:
To install the shell, you will need to send a memory backup file. Note that this replaces the entire calculator memory contents. First, put the calculator into link mode if necessary:
Next, use the Restore option in TiLP to send the .73B, .82B, or .85B file to your calculator. You will need to press [ENTER] on the TI-82, or [F1] on the TI-85, to accept the memory backup.
If you have a TI-82 or TI-85, put the calculator into link mode again (using the same key sequence as above.) Next, double-click on the Operating System item in TiLP. It will show a warning message:
It will ask you again to confirm that you know what you’re doing:
The ROM dumper will then be transferred, and if possible, launched automatically. On the TI-73, TI-82, or TI-85, you will need to launch the ROM dumper manually:
The ROM transfer will take some time…
In addition to emulate the behaviour of a real calc, TilEm2 features main
functionnalities as file loading and file export, screenshot, debugger.
In this chapter, we will talk about these main features.
In an another chapter, yoou will find a detailled explanation of all options of
TilEm2.
Here we talk about sending a file from computer to emulated calc.
You downloaded a file on the web and you want to test it before sending it to your
calc?
You compiled a file and you want to see the result?
So you probably want to use send file feature...
There’s 3 ways to do that.
Firstly right click on the calc.
This menu popup will appear :
Click on Send file...
(You can bypass the popup menu by clicking CTRL + O)
Then explore your computer and choose the file(s) you want to send.
Usually, variable are suffixed by something like .82p (ti82), .83p (ti83) .8xp (ti83+,
ti84+), .86p (ti86) or something else.
Grouped files are generally suffixed by .82g (ti82), .83g (ti83), .8xg (ti83+, ti84+).86g
(ti86) or something else.
Some other special extension as .8kv are flashapp for ti83+, ti84+.
And a lot of other file extension.
It could take some time to load a variable so a current progress bar is printed
while loading to know what’s happening.
As soon as you press OK, the file start to be loaded and a progress bar is
displayed.
Simply select one or more files on your computer and use drag and drop to
You will not see any visual feedback (no progress bar) but you can see in your
terminal eventually the libticalcs debugging messages.
You can check if your program is correctly uploaded to calc by listing them inside the program menu (if it’s a program) :
You can also send a file to the calc at startup using command line parameters.
All the non options args are sent to the calc.
This example will load the group file ion.83g in the calc memory.
To get a program, list, screen, application or whatever which is considered are a var
on the calc you need to use the right click menu options :
(Or you can simply use CTRL + S)
At this point, TilEm2 will (try to) get the variables and print them into a
list.
This is why you will see the progress bar.
Warning : update is not done each time the window is popup, you need to refresh
manually the vars.
You can select one or more vars in the list then saving it by clicking ”Save”.
A check button lets you choose the format of the output.
If you want to get more than one var, you can save it as grouped file or as separate
files (separate files is as you get it one by one).
Then simply click save and choose a place to backup the var(s).
As screenies are a good way to show your projects or whatever you want, TilEm2
provide a complete screenshot dialog.
We think this dialog to be user friendly and powerful.
There’s also a special manner to quickly screenshoot the lcd content (”Quick
Screenshot” or SHIFT + CTRL + PRINT).
We will describe this method in a first subsection then we will talk about more
powerful methods.
You can grab a screenshot without using screenshot dialog.
Simply click on the ”Quick Screenshot” menu option or SHIFT + CTRL +
PRINT.
The screenshot use the default options (or the options you have given into screenshot
dialog).
The picture is stored into the directory you usually use for screenshot OR if not
exists into the /.config/tilem2/screenshots/ or equivalent if you’re not on
Linux.
Click on ”Screenshot” menu option or use CTRL + PRINT.
The screenshot window will open.
TilEm2 automatically grab a screenshot at startup.
You can choose to ”Save” it or to grab another screenshot.
There’s a lot of options as size, foreground/background color etc...
There’s some differents kind of output format as png, bmp or some other
else.
Click on ”Screenshot” menu option or use CTRL + PRINT.
The screenshot window will open.
You can record a gif by clicking ”Record”.
Stop the gif by clicking ”Stop” (What a surprise :D).
As soon you click ”Stop”, a preview is available in the picture area.
There’s a lot of options as size, foreground/background color etc...
You can set these options after recording the animation.
When you have the desired animation, clickon ”Save” button and choose a place
and a filename to save the gif.
Static screenshots and animated screenshots both use settings. Here’s how to use
these seetings.
Size
You can choose between ”default” size for the screenies :
7 default size for the TI-86 and 3 for the other models.
For the TI-86, ratio and values proposed by default are different:
Default values are proposed to help you to choose correct ratio (the choice
depends on the size of the lcd).
Here’s what you have for all models except TI-86 :
288 x 192
I will not show the same screenshot for the TI-86 it’s not necessary.
This is ”default” sized screenies but you can set your own (warning : try to keep
the ratio).
Setting your own size could produce some curious results (here the biggest possible)
:
Animation speed
There’s a setting called ”Animation speed” which is used by the animated
screenies.
If you increment this value, the animation will run faster.
A visual feedback will show you the effect of increasing this value.
The max is set to 100.
Foreground and background colors
Below the ”animation speed” there’s 2 color chooser which allow you to set your
own colors for foreground and background colors (as you know z80 calc LCD is
monochrome).
The grayscales screenies are fully compatible with this setting of course (because
it’s no more than ”flashing”).
Here’s a sample of what you get if you choose a pink as foreground color and green as
background color :
Let’s see a bunch of colored screenies:
Of course these settings are usually used to correct contrast or simply set a
better color ratio (not to do the useless but funny screenshots I’ve shown just
before).
A last setting called ”Grayscale” let you choose the rendering mode.
This is an important part of an emulator...
The debugger provide an easy way to inspect the core of the calc and to find bugs
into your projects.
When you start the debugger (using ”Pause” or the right click menu entry)
The calc will be paused automatically (you can’t work with debugger while calc is
running that’s evident).
If you click on F5 to run the calc the emulator will be deactivated to prevent to edit
anything in the debugger.
You can pause the calc by clicking ESC (escape).
Or by using the ”Pause” button (tool bar behind the menu bar).
The aim of the debugger is to show the disasm memory, the registers, stack and
the memory.
The values are written in hexadecimal format.Not decimal.
What does it mean?
Simply one little sample :
increment 0 = 1;
increment 1 = 2;
increment 2 = 3;
increment 3 = 4;
increment 4 = 5;
increment 5 = 6;
increment 6 = 7;
increment 7 = 8;
increment 8 = 9;
increment 9 = A;
increment A = B;
increment B = C;
increment C = D;
increment D = E;
increment E = F;
increment F = 10;
Ok we will not explain more the hexadecimal format, just accept the fact that’s in
hexadecimal format.
For a large part of TilEm2 users, hexadecimal is not a surprise :)
Let’s talk about widget organization.
There’s at least 6 zones into this window :
The disasm view :
The memory view :
The register view :
The stack view :
All this view produce the debugger (there’s some other window hidden by
example breakpoints dialog and keypad window).
As you probably know, there’s a register which store the adress of the current
instruction.
It’s called PC (”Program Counter”).
If you look into the value of PC, you should recognize that the value is the same as
the adress pointed by a yellow arrow in the disasm.
This is the current instruction which is executed.
See this example :
As you may know, even if z80 is most like ”RISC” methodology (in the sense the
mnemo are pretty simple and fast) the size of the instruction depends the instruction
itself.
That’s why the adress are not linearly incremented.
Sometimes an instruction is stored on 1 byte, sometimes 2 bytes, sometimes 3 bytes
etc...
A sample for 1 byte sized instruction :
A sample for 2 bytes sized instruction :
A sample for 3 bytes sized instruction :
This view show the disasm memory, so you can see the entire memory, but not as
hexadecimal value but as assembly mnemos.
As you can see, some values are replaced by their symbols.
This is not a default behavious of a z80 debugger, so that’s the job of
an equate file loaded by TilEm2 at startup (files called .sym into the data
directory).
In fact that’s just adress values replaced by a label (for human readable
purpose).
Exactly as you use an equate file instead of calling directly the adress for a
call.
When you use ”call BUFCOPY” (ti83), you don’t want to see the exact value of this
jump.
Always about the disasm view, you can ”Goto adress...” :
Using right click on the disasm view :
Or using the ”Go” menu into the menu bar :
This option lets you choose an adress to ”jump” (only user interface no action on
the calc).
Because it can be very annoying to scroll the disasm view.
One more time : ”Goto adress” doesn’t change anything to the ”PC” or
whatever.
You can also use CTRL + L to do this task without using the mouse.
Another option is to jump to the ”PC”. It’s like goto adress, but without
prompting an adress, it will jump to the ”PC” directly.
Using it from the menu bar (”Go” menu) :
Or right click on the disasm view :
Now let’s talk about interactivity.
As you can see, a a lot of informations are updated each time the calc run (or do a
step).
Stack, registers, and memory is updated in the same time the disasm view is
modified.
There’s of course some possible actions to do to execute one or more steps.
Here we will only study disasm related stuff, keeping breakpoints and registers/stack
for later.
The actions you could do :
Step
You can execute this action by clicking on the button in the button bar
:
Or click on the ”Step” menu entry into the ”Debug” menu into the menu bar
:
Or simply use ”F7”.
This action simply execute one step.
Warning : Execute one step doesn’t mean going to the next line !
If the instruction is a call or a jump, it will load another value into the ”PC” so
possibly jump elsewhere.
This action is the best choice to follow ”step by step” the behaviour of a
program.
Step Over
You can execute this action by clicking on the button in the button bar
:
Or click on the ”Step Over” menu entry into the ”Debug” menu into the menu
bar :
Or simply use ”F8”.
This instruction is useful to do not enter into subroutines.
What’s a subroutine?
Simply a jump materialized by a CALL (or BCALL) which push the return adress
and finish by a RET.
Finish Subroutine
You can execute this action by clicking on the button in the button bar
:
Or click on the ”Finish Subroutine” menu entry into the ”Debug” menu into the
menu bar :
Or simply use ”F9”.
This action simply run to the RET instruction then pause after the execution of
the RET (which is basically poping the value on the stack into ”PC”).
This ”Finish Subroutine” is helpful if you enter a subroutine which is quite long and
you don’t want to inspect it, so simply finish the suroutine easily.
In addition to run step by step and inspect the disasm instruction, you would
probably know what’s happening when you run one instruction.
Why a jump is never executed? What’s the content of a register after an instruction?
What’s the flag state after an instruction?
The register view provide an easy way to inspect and edit the register values.
Firstly some asm z80 reminiscence :
The values of these registers are updated (not always) and you can inspect what
your program is doing.
You can also edit these registers/values just by clicking on the input and replace the
current value.
There’s a bunch of toggle button to show and edit the ”F” flag.
If you click on a button, you will set or reset a bit of the ”F” register.
You can see that each time you toggle a button the ”AF” register is modified.
You can change the values of the register as you want for testing what could
happen if a value was different or whatever you want.
If you try to change ”PC” value, you will see the disasm view updated.
Warning : this is usually not useful but why not...
If you try to change the ”SP” value, you will see the stack view updated.
Try by example to set ”FFFF” as value of SP... Then the stack view only contains
one value !
TilEm2 provide a view of the stack represented as a list of ”address: value”.
You can navigate through the values in the stack to inspect them.
Let’s a little explanation about what’s a stack?
The stack is a particular memory access to store an retrieve values using the LIFO
method.
The higher place in the stack is always the lower adress.
For ti83, the stack starts at adress FFFF for the first value.
Each time you PUSH a value, the ”SP” is decremented by 2 (each place is 2
bytes).
A POP increment the ”SP” by 2.
As said previously, if you try to set manually the ”SP” to the value FFFF
(for ti83), you will completely drop all values of the stack except the last
one.
This view is basically more a feedback because you can’t edit it.
There’s an interesting feature in TilEm2 which help to find the instruction which
modify the stack.
You can find these actions in ”Go” menu into the menu bar.
Or using ALT+PAGEUP or ALT+PAGEDOWN.
When you use this, you will jump (not really jump) to the instruction which
modify the current stack value.
It will update your disasm view but change nothing to the current state of the calc
(no change in registers, ”PC” not modified etc...).
This should be easy to understand that this view only print the memory in
hexadecimal format.
You can scroll through the memory and edit it (when it possible).
To edit, simply click on a particular value.
Now, try to edit the value in the adress 0000 (for ti83).
You simply can’t !
Why TilEm2 do that?
Simply this is a place of ROM (so read only).
In fact you can’t edit the entire memory but only some part (but your programs and
variables will always be in RAM so you could edit it without any problem.
Now that we have seen a big part of the debugger, let’s see an option which will
change the notation of the adress into logical or absolute representation.
Here’s the result :
There’s a completely independant widget which is poped up when you click on
”Keypad” into the ”View” menu into the menu bar.
This widget allow you to see how the buttons are connected (which bit in which
group).
This is mostly used in direct input (use non blocking port to scan keys instead of a
system call as CALL GETKEY).
Imagine that you want to inspect a part of code, but how to run the program
normally and stop just before the part of the code you want inspect?
Try to run/pause randomly to finally stop in the right zone is stupid (and hard to do
usually).
The solution is to use breakpoints.
What’s a breakpoint?
A breakpoint is just say ”When the ”PC” is equal to this adress, then pause the
program”.
Assuming you have a routine called ”DrawGbuf” and you want to stop just at the
start of this routine.
How to know the adress?
Simply use the correct option in your assembler to generate the ”listing”
file which contains the equates between labels and adress (-T option with
spasm).
Then look for the ”DrawGbuf” label into the file (with spasm the file use the ”.lst”
extension).
I found the adress 9e4f (hexadecimal of course).
As you can see, a label does not take place into the final binary, it’s only a name
for an adress only for human.
Now you have the adress, simply run TilEm2, launch debugger, then open the
breakpoint dialog :
In the menu bar, click on ”Debug” and click on ”Breakpoints” :
Or simply press CTRL+B.
Then the ”Breakpoints” dialog appears :
As we want to set a breakpoint on 9e4f (which is the start of the ”DrawGbuf”
subroutine), we need to ”Add”, so simply click on ”Add”.
Another dialog opens and we simply keep the default values and type 9e4f in the
”Adress” input.
The list of breakpoints is updated and as we can see the breakpoints is
present.
Now use ”Run” from the button bar, of the ”Debug” menu or simply
”F5”.
The debugger will stop when the adress is encountered (could be never).
In our case, I know that ”DrawGbuf” is frequently used because I use it as soon as
possible to refresh the lcd.
So as soon as I do ”F5” (run), the debugger stop on the 9e4f adress.
As my first idea was to inspect the instructions into this subroutine, I can step by
step and eventually do ”F5” which will run and breaks immediately (I’ve said this is
used very often in my code).
Ok I assume you have understood this basic example of uses breakpoints.
Now, I want to add another breakpoint on 9e8f which is a bit higher in
memory.
So CTRL+B then ”Add” then put 9e8f into the text input.
Now I have :
Finally I see it’s was never reached so I want to change the adress so I launch the
breakpoints dialog and click ”Edit”.
And I put an adress just after the first breakpoint.
Then I run, and I see that the second breakpoint is reached :).
Now If I want to delete a breakpoint, I simply use ”Delete”.
If I want to keep the breakpoint but just deactivate it temporarly, then uncheck the
checkbox into the ”Type” column.
I can break the program using a range of adress.
By example, I want to stop the execution between 9e4f and 9e56 because I’m not sure
where exactly to add the breakpoint.
So I deactivate my previous breakpoints and ”Add” a breakpoint with a range
of value (use toggle button ”Range”) starting from 9e4f then finishing at
9e56.
Our list look like this now :
And each instruction inside this range is marked as ”Break” :
Now, I want to put a breakpoint on each RET.
So I use ”Add” then I choose ”Z80 Instruction” into the ”Breakpoint type”
listbox.
As you know RET is ”C9” as opcode so just type C9
Then we have :
Ok now just remove this breakpoint.
You can also set a breakpoint when accessing the ports.
As I use ”Direct Input” method to scan keys in the project, I want to see when I read
from the port 1 (keyboard).
So I click on ”Add” then I choose ”I/O” port then I check ”Reading” then I give the
01 value for the keyboard port.
Let’s see if all is correct.
And of course it works fine :
As you can see it stops just after reading the port (on the CP mnemo which is the
group to test).
Now I want to stop on ”Undocumented instruction” and ”Illegal instruction”.
An undocumented instruction is an instruction which was not explained into the z80
datasheet but usable anyway.
A illegal instruction is just a bad instruction .
To do that, I simply check the check button like this :
Finally, as I finished to debug my application and just want to run it normally, I
just clear all the breakpoints by clicking ”Clear”.
Simply click ”Clear” then all will disappear.
To finish about breakpoints, you should know that you can check ”Reading”,
”Writing” and ”Executing” for a ”Memory adress” breakpoint too.
You can also specify logical or absolute adresses.
You can set a breakpoint easily without using the breakpoint dialog by doing
right click then ”Breakpoint Here”.
It will put a breakpoint on the current instruction selected into the disasm
view.
This chapter is most like a dictionnary than real explanation. Please refer to the first
part (”Basic tasks”) for the subject which are already explained.
As TilEm1, the menu is a popup menu (right click).
All you want to do need to use this menu.
As you can see, there’s all you need, no more, no less:
This is one very important feature, because emulators are usually used to try some
programs before really transferring it to real calc.
When you click on this menu entry, a file chooser dialog is opened and let you choose
a file.
A lot of people don’t know which file extension is associated with the emulated
model...
To help them, some patterns are used to do the selection.
When you let ”All compatible files”, TilEm2 do the job for you, but you can choose
”All files” if you know what you’re doing (a file with an incorrect extension by
example).
It could take some time to load a variable so a current progress bar is printed
while loading to know what’s happening.
When you click on ”Receive File...” menu entry, TilEm2 firstly get the vars then
prints it into a listview.
After the first launch, refresh is made only on request !
If you click ”Receive File...” then close the window, then create a program and click
”Receive File...” you will not see your program.
The variable list let you choose the stuff you want to backup.
If more than one variable is selected, you can choose between two modes of
backup : ”Separate files” or ”Group file”.
If you choose separate, each file is saved as if you have saved one by one.
If you save grouped, a group file will be created on disk.
When you finally click on ”Save” button, a file save dialog is opened.
Choose a directory and a name and click ”Save”.
When you click on ”Open Calculator...”, a file chooser dialog pop up and let you
choose a rom to load.
So in fact even if you already emulates a calculator, you can switch to another just by
opening a new rom file.
Another way to do that is to quit TilEm2 and restart it using another rom file
(option -r).
Rom files usually finish by .rom as extension but you can use ”All files” pattern if
you have a rom with a odd extension.
This option just save the current state of the calculator in a .sav file.
The file is created in the same directory as the rom file and with the same
name.
No surprise, this option just revert the calculator state (if possible).
Guess what does this option :)
When you click on this option, the debugger window will appear.
There’s a lot of things to say about debugger.
When you launch it, calculator is automatically paused.
As you can see there are 5 big buttons : Run, Pause, Step, Step Over, Finish
Subroutine.
Step just execute one instruction.
As you can see, all the instructions are not the same length, that’s why it doesn’t
step one byte per one byte.
Step over do the same job than step but do not follow call.
Finish subroutine just do basically the same job but stop after a ret.
Now just see what’s the differents view of the debugger dialog.
There’s a big frame for disassembly view.
In this frame, you can see the adress and the disassembly instruction.
On right click, you can do some useful actions : Breakpoint here, Go to adress, go to
PC.
There’s 2 kind of adress notation for this view : Logical and Absolute.
You can switch it into the ”View” menu.
The second big frame is the memory view.
For this view you can switch the adresses representation if you want.
In this view you can see what your calculator contains.
You can also edit the memory and change some values by your own.
A third view represents the registers.
You can edit them too.
Below registers there is a bunch of toggle button to represent the flags (you can
change it).
Then Interruption Mode IM, I, and Enable Interrupt (checkbox).
The finally the stack.
At the top of the debugger window, you can see a menu ”Debug”.
The options are the same than buttons but there is a big news : breakpoints.
Breakpoints are a big part of the life of a assembly developpers.
This option opens the Breakpoint menu when you can ”Add”, ”Remove”, ”Edit”,
”Clear” or some special action like ”Break on invalid instructions” or ”Break on
undocumented instructions”.
The easiest way to add a breakpoint is to right click on the disasm view and click
on ”Breakpoint here” but you can also set a breakpoint using its adress (logical or
absolute).
The ”Go” menu basically provide and easy way to navigate into the disasm view
and th e stack.
Now I must talk about a nice tool called Keypad (into View menu).
Macros are an easy way to simulates key press, file loading, reset automatically.
It means that you could record a macro then click on some keys, then stop.
If you play it, tilem will press the same keys for you.
Have you never think too lazy to press always ”2nd catalog asm(” each time you
want to test your new asm production.
Simply use a macro to load and launch your program automatically !
About the options, you can play an already loaded macro or a macro you just
have recorded.
You can also open a macro and save the current macro (which one you just have
recorded).
By clicking on this option, you launch a screenshot dialog.
You can grab static screenshot (multiple format) or animated screenshot (will be
saved as gif).
As you can see TilEm2 has a lot of screenshot configuration.
So you can change the size, change the foreground and background colors.
Use or not grayscale.
By clicking on this option, you grab a screenshot.
This is where you can set the skin (or disable using it) and some important other
stuff.
You can limit speed or not.
Emulate grayscale (if you don’t know just let it checked by default).
Use smooth scrolling.
And an important user friendly feature...
Set skin !
When you click on the button, a file choose will popup and lt you choose the
skin.
No more than an about dialog :)
Bye Bye ;)
TilEm2 is basically made for Linux and command line is our first love :)
Here the options at launch.
You should usually use something like :
All the non option arguments are considered as files to be loaded.
But as you can see you can specify the skin with -k.
If you usually use more than one model, you can try -m and it will load
the rom associated with this model (if you already start a rom from this
model).
You can specify a different save state (by default it uses the one which is called as the
rom file).
You can start skinless.
You can load a file and even launch a macro at startup (in this case loading a file is
done before macro playing).
You can reset too, get a var (if possible) and launch debugger.
Some options could be set at startup as normal speed or full speed (as fast as
possible).
Other options are not TilEm2 options (–display by example).
Something is missing?
First starting using configuration file (need to have already started one time before).
You can specify the rom to use:
You can choose a skin at startup
Choose a save save state:
Or just starting a model without giving a rom file:
Starting skinless :
Reset the calc at startup :
Trying to get a var at startup:
Starting in full speed mode:
Or in normal speed:
Play a macro at startup:
Warning, in this situation, load + macro play, the load is done before the macro
playing.
You can also launch debugger (and pause the calc by extension) :
You should not edit this file because you don’t need it.
If something wrong, by your fault or a bug in TilEm2, just get a new configuration
file from the TilEm2 website and replace the wrong config.ini file.
It’s usually copied into /.config/tilem2/config.ini
The keybindings are defined in a keybindings.ini file (usually copied into
/.config/tilem2/keybindings.ini).
There’s currently no tool to interactively edit this file, but you can edit it by hand if
you do this carefully.
This file uses inheritance, it means that a part of the keybindings are common to all
models, but you can rewrite them for each model.
So firstly, TilEm2 parses common, then if he find another keybindings for the same
keypress, he rewrite it.
There’s one subsection per model.
Here’s a piece of the keybindings file :
Left values are the computer keyboard key names.
Right values are calc keys.
+ means pressing the two or more keys simultaneously.
, means the same thing for the calc.
We know that it’s not easy for you to modify this file, but if you make a mistake,
tilem will say ”error while parsing keybindings” but it works even.
In the future, we will probably add a user interface integrated to TilEm to change
keybindings, but for the moment it doesn’t exist !
Here are some tips to help you to develop your projects.
Each time you compile your program, if you must do 5 or 6 click to launch you
program on the emulator that’s pretty annoying.
Record a macro to do that and execute it at startup.
If you load files, they are stored into the memory before the macro is launched.
So record a macro and put it with your project (here into macro/play.txt).
Here’s a sample of what you can do:
I put this code into exec.sh then for testing my program I simply do
:
And my file is loaded and launched.
As mentionned previously, TilEm2 uses a TiEmu skin file format.
So it’s easy to do your own skins.
Skinedit is a part of TiEmu, you can download it here :
http://www.ticalc.org/archives/files/fileinfo/232/23201.html
Now, I will explain to you how to create your own skins.
A picture of a game boy will be our example.
Firstly you need to download tiemu-skinedit (author : Julien Solignac).
Then launch it by :
Or by clicking on the icon if you have it on your desktop.
skinedit starts...
Click on File -¿ New .
Choose a picture then click ”Ok”.
Then find a picture and resize it to 900 pixels of height (keeping the
proportions).
Usually I use TheGimp like this.
But it works with any size, just 900 pixels is nice I think.
The picture is loaded and a popup ask you to give some informations (title, author,
model, contrast).
The title is a comment added in the skin file.
Same thing for the author which is very important for licencing informations.
Model is the model of the calc (here I used ti83.
Contrast is the contrast for the lcd, you can choose the background and foreground
color (or just set ”low contrast” or ”high contrast”).
Then click on ”LCD” then select the region dedicated to LCD.
When you’re satisfied by the region, simply do a right click to save the region.
Then click ”keys” button.
Select the key, then define a region (select a region then right click to
save).
What I could say about the skin format is that each key is defined by a
rectangular region (there’s a max number of keys hardcoded at the value
60).
So each time you define a key, it will be saved as a rectangle into the skin file, the
other unused slots keep the 0,0,0,0 values).
LCD region use the same system, with some more information (contrast...).
When you’ve defined all the keys (usually it’s what you should do) or only some of
them (as it’s the case in our funny example), you can check it one by one to be sure
it’s ok.
The keys are listed in the right order. First key in the list is the left top key (usually
[Y=]), last key is the right down key (usually [ENTER]).
Warning : do not supperpose regions (it’s sometimes hard to do with arrow
keys).
For this tutorial, I just defined arrows, start = [ON], select = [2nd], B button =
[PRGM] and A button = [ENTER] When it’s ok, save the file with a ”.skn” file
extension.
And let’s see the result !
You can also convert a TiEmu or VTI skin file using skinedit then use it with
TilEm2 !
Do not hesitate to send us your skins file, maybe we could use it as official skin
!