How Do I ......????
Ben Duncan
bns
Mon Oct 11 09:30:42 PDT 2004
First, thanks for all the good advice, now ....
The GLOBAL variables need to be GLOBAL. The premise is that they
are truly dynamic and any function has the right to change them.
This is limited to the variables that are affecting the screen
row, column and size, as well as data entry fields. They are defined
as extern in the header, and allocated in the slag-module, which is
the "driving" module of the whole thing. This is also done so that in a
case of a "trigger" key event, the call back module will know of
all the right data as well as positional information
integrity flowing smoothly.
In the case of the S-Lang library, it is also imperative to pass by
reference and not by value, information to S-Lang. This is because
the information needs to available both to the S-Lang functions as
well as the interpreter. In my case, the actual "slag" kernel will
probably be written in slang utilizing the "driver" for speed.
FWIW, at this time, I am able to populate "root" screens and "window"
screens, do data entry to both types, put up "text" and "selection"
boxes (although I have not done the data retrieval from selection/radio
boxes routine yet). I can maintaine screen display integrity while all
this happens. I have shadowing on windows / boxes with 3 effects, a
"System header" (First row), a status line (last row), system information
line (1 up from last row) all working and maintaining integrity without
getting mangled, or stepping all over each other. I have selectable
scrollable data entry (Limited to alphanumeric only), complete
function key / keypad recognition working (they don't do any thing
at the time, just a message saying which one has been pressed) all
working.
Granted, It is very rough right now. I still have menuing, file input/output,
entry "edit" maps, and the "language" of the SDEF's (S-lang definition files)
to work on, but I have come a pretty good ways already.
Seems I have been learned a tremendous amount about C since I posted
that message.
Rick Sivernell wrote:
> On Sat, 9 Oct 2004 08:22:10 -0400
> Kurt Wall <kwall at kurtwerks.com> wrote:
>
>
<SNIP>
--
Ben Duncan - VersAccounting Software LLC 336 Elton Road Jackson MS, 39212
"Never attribute to malice, that which can be adequately explained by stupidity"
- Hanlon's Razor
More information about the Linux-users
mailing list