Wow. What a weekend! I'm bowled over by your progress Scotty and others.

I have made these changes to /mnt/settings/er_registry.txt

1. add ;SHELL to the end of the line that starts uaIDList

2. add these lines under the EBA section:

[SHELL]
category=viewer
arguments=<FILENAME>
exec=/bin/sh
xResourceName=sh
channel=3
extensions=sh

... then reboot.

Once done, you will be able to execute shell scripts, including the network and dropbear startups by clicking on them.

Note that /mnt/settings/er_registry.txt is occasionally overwritten from /etc/er_registry.txt, so you may want to change that too.

Once done, you will be able to execute shell scripts, including the network and dropbear startups by clicking on them.

A worthy hack!

Time to find an xterm for the iLiad so I can write a script to start it. :)

Time to find an xterm for the iLiad so I can write a script to start it. :)

I think the ncurses and rxvt worked in 2.4, they should work now:

http://iliad.mobileread.com/feed/

In theory rxvt has small footprint than xterm. Xterm is a hell to compile in cross environment because you have a lot of options for your ncurses (appropiate name, it has).

If you install the native toolchain you no longer need to cross compile. :D

This is terribly exciting ... way over my head, but exciting! :grin:

So let me ask the dumb question: does this mean we're getting close to being able to install & run Linux versions of other reading apps?

This is terribly exciting ... way over my head, but exciting! :grin:

So let me ask the dumb question: does this mean we're getting close to being able to install & run Linux versions of other reading apps?

Pretty near. As of today, we could do a reading app using the pen as input. A different matter is to use the keys in the iLiad; we should need either the library headers of the promised SDK or to build our own headers to the libraries from reverse engineering, ie from using ltrace and gdb on the e-ink applications.

you are not only close....

But i think what lacks most applications is the support for driving the e-ink panel, especially when it comes to partial updates. And that would mean you have to recompile the app with calls to a library that refreshes the screen.

Have I got that together right?

you are not only close....

But i think what lacks most applications is the support for driving the e-ink panel, especially when it comes to partial updates. And that would mean you have to recompile the app with calls to a library that refreshes the screen.

Have I got that together right?

Yes, but if gdb is working we can imagine some headers for the liberdm.so.2 by tracking the calls between scribble and this library. Keryboard is to be more obscure because it seems to go across the iLiad messaging architecture, whatever it is.