Mark Wedel wrote: > I don't see any problem with the first one. Committed to CVS. > I'd have to check the CVS logs, but there is probably a reason that there > is an explict check for FLAG_REMOVED in the second case. My guess would be > you could ge in odd situations with a logged out player, possibly resulting > in crashes. I checked some random code with loops over the first_player list: I found no other player that checked for FLAG_REMOVED or even bothered to check "pl->ob!=NULL". Therefore you can cause the problem fixed in find_player() ('reply' command) by just using 'tell' (find_player_partial_name()) or 'gsay' (send_party_message()). > In fact, I've seen a lot of crashes with logged out but still connected > players. Really, all that logic should get cleaned up/removed (I think I > mentioned before, but ideal situation would be for client to send both name > and password as part of connection instead of having the the server have to > handle those states, which I think there is an open bug against right now) I'll leave it unchanged then. Anyways, it was just a little inconsistency I noticed between the similar commands 'tell' and 'reply'. _______________________________________________ crossfire-devel mailing list crossfire-devel at lists.real-time.com https://mailman.real-time.com/mailman/listinfo/crossfire-devel