Re: JOURNALING - the quest for the dox goes on

[jsimmons@cs.sunysb.edu (Jason Simmons)]

In article <1993May8.175526.16803@ifi.unizh.ch> gudat@avalon.physik.unizh.ch (gudat henrik) writes:
>Maybe we can find out together how we get that done (Jason - are you
>listening??:)

I hope we can find the necessary information about journaling, since it
would definitely be useful for a macro utility.  For the time being, I have
been concentrating on a programming-langauge approach instead of a macro
recorder.

>- One of the possible calls to the journaling driver is GetNextEvent.
>However, the event record hasn't a fixed length - it can be the 'old'
>record or the extended one. How many bytes have to be preserved?

Since journaling is part of the Event Manager and can be used without the
Window Manager, I think the GetNextEvent data would only be an "old"
event record (16 bytes).

>- When playing back a sequence, I put the data into the memory
>locations that start at resultPtr. Is that correct?

I'm not sure.  ResultPtr looks like a temporary storage area for the event
information.  If the system was rebooted between recording and playback,
ResultPtr would probably point to some random area in memory.

>- now the most important question: HOW DOES PLAYBACK WORK?
>The Event Manager returns JournalFlag=record and resultPtr=pointer to a
>data structure - but what about the JournalCode? If this word is
>meaningful, what happens when JournalCode does not match the current
>packet? Just advance (in the macro sequence) until one of this kind
>appears? If JournalCode does not contain anything of interest, how does
>the Event Manager know what action comes next?

The JournalCode reports which Event Manager call was mode.  The code values
are listed on page 7-20 of TB Ref I.  I don't know how the journaling
information recorded by the driver can be replayed.  I posted this question
a couple of times previously, but nobody seemed to have an answer.

By the way, page 31-2 of TB Ref III has some information on a sixth journal
code for ReadMouse calls.

Jason Simmons
jsimmons@sbcs.sunysb.edu