[Asterisk-Dev] quick question about the new args parsing stuff...

[Chris Wade]

[* ducks behind 'firewall' just in case *]

Just wondering, even though its probably already been discussed on the 
dev conf call, but why is - as an example - app_skel implemented this way:

static int app_exec(struct ast_channel *chan, void *data)
<snip lots of code using some arg splitting/processing functions still 
replicating functionality from one app_exec to another - code that 
builds an int argc and char **argv>

would it not be better to do this:

static int app_exec(struct ast_channel *chan, int argc, char **argv)
<snip much less code that everyone knows how to write, ie. 'int main'>

I thought the point of the whole args parsing stuff was to have the API 
handle the grunt work and leave the programmer to the new innovative 
stuff.  I could be wrong :^)

Now as for what I seem to remember being argued on the dev conf call 
about this being such a 'fundamental' change that is not backwards 
compatible, why not create a new app registry and begin letting us 
newbie grunts start porting old apps to the new registry and style? 
HINT-HINT:  I'm volunteering my time for the porting of existing apps to 
the new registry, but I cannot write the registry itself - not that good 
at C and/or * internals.

-Chris