> I have made a set of changes to the consoles such that F1 is now the 'native'
> console.
Hey, good stuff! I'd been thinking about this but coudn't come up
with an elegant way to do it before I had to move on. I'll be interested
in seeing your code. (thanks!)
> My next step is to increase the size of the buffer and somehow figure
> out a way to be able to page up and down on the native console so that
> we can view output that has scrolled out of view.
As long as you're adding scrolling/buffering to the native
console, feel free to add it to the virtual consoles, as well :)
> In doing this I also discovered a bug with the decaf native methods. I had
> made a new class, consoleNative, which implements ConsoleDriver just like
> jos.platform.drivers.console does.
BTW -- having worked with this way of doing the drivers/consoles,
do you have any structural-style suggestions?
> In consoleNative, putChar and setChar became native methods.
> Suddenly, all keyboard input was outputted to the native console
> instead of the java consoles. What was happenning is that the native
> consoleNative.putChar() was being called where console.putChar() was
> supposed to be called. I changed consoleNative.putChar() to instead
> call the native consoleNative.putNativeChar() and suddenly everything
> worked. In short, there is a problem with how decaf resolves if a
> native method is supposed to be used. I haven't tracked down what
> exactly is happenning yet, but I did leave the debugging messages I
> used to track this down in the code but commented out along with what
> to do in order to show the bug again so that someone can track this
> down later.
That is a truly twisted bug, if I understand what you're saying
correctly. You've got N consoles, where the first one is
driver.consoleNative. For consoles 1 to N, which are driver.console
instances. When 1 to N .putChar() is called, decaf calls the
driver.consoleNative.putChar()? -- That is, decaf called not only the
wrong type of method, but also a method in the wrong class?
> P.S. I was going to include a patch file with this, but the cvs server is down
> again. Could someone with access please restart it? I'll send the patch as
> soon as I see that cvs is up again. - Thanks.
I turned it back on.
-_Quinn
_______________________________________________
Kernel maillist - [EMAIL PROTECTED]
http://jos.org/mailman/listinfo/kernel
