store.c store_maint function

[PaulBlay]

I noticed the following line in store_maint

while (st_ptr->stock_num < stock) store_create_random(which);

This does not allow a way for store_create_random to fail in a helpful manner and can lead to an endless loop freeze. (If somebody like me has been mucking about with the game code ;-)

How about something with a for loop and a maximum number of tries?

Could also return values from store_create_random and/or store_maint to show if the process has gone as planned or not.

[PaulBlay]

I wrote some suggested alternative code, but there's something wrong with it So more debugging for me.

[takkaria]

I noticed the following line in store_maint

while (st_ptr->stock_num < stock) store_create_random(which);

This does not allow a way for store_create_random to fail in a helpful manner and can lead to an endless loop freeze. (If somebody like me has been mucking about with the game code ;-)

How about something with a for look and a maximum number of tries?

Could also return values from store_create_random and/or store_maint to show if the process has gone as planned or not.

A simple change would be to do:

while ((st_ptr->stock_num < stock) && store_create_random(which)) ;

and make store_create_random() return FALSE when it fails, TRUE otherwise.

[PaulBlay]

make store_create_random() return FALSE when it fails, TRUE otherwise.

Waddya know. store_create_random already does return FALSE when it fails, TRUE otherwise.

However I suspect it is possible for the functions called by store_create_random to fail (like store_carry) which leaves you with a created object (success) that hasn't increased the stock number. That's another way that you can end up with a potential for infinite loop. Failing on a failed store_carry also would be dubious as just because one object doesn't fit may not mean that the next one won't go in just fine.

I shall obviously have to consider this a little more.