Talk:GTS sabotaged by greed: Difference between revisions

Ideas for fixing
(Ideas for fixing)
Line 79: Line 79:
   
   
I got those options while looking for Chimchar too,but I also found some reasonable trades. i got a level four Lopunny in exchange for my Gabite (it said ANY level), A Stunky in exchange for a level 9 & under Muma (mine was lv. 7) & ANY Piplup for a Granbull( mine was level 1, freshly hatched). The only problem I encounterd is you can only search for pokemon you have seen before (so if you want a pokemon that you haven't seen you can't get it untill you do. This makes me mad because I wanted a female Sedot... [[User:Gatogirl|Gatogirl]]
I got those options while looking for Chimchar too,but I also found some reasonable trades. i got a level four Lopunny in exchange for my Gabite (it said ANY level), A Stunky in exchange for a level 9 & under Muma (mine was lv. 7) & ANY Piplup for a Granbull( mine was level 1, freshly hatched). The only problem I encounterd is you can only search for pokemon you have seen before (so if you want a pokemon that you haven't seen you can't get it untill you do. This makes me mad because I wanted a female Sedot... [[User:Gatogirl|Gatogirl]]
== Ideas for fixing ==
Okey, first off I'm just brainstorming ways to fix a half-broken system here.  I don't expect that any of this would ever be implemented, as I don't think the programmers designed Pokemon D/P to be patchable an the client side...
I've found that putting a pokemon up for offer works amazingly well (if you're making a reasonable request), but I've yet to find a viable trade in the search.  For that matter, all I ever see in search are dupe-bug abusers (asking for impossible pokemon).  First off, how to fix the dupe bug:
* Validate the requested pokemon on the server-side.  It wouldn't be too hard to deny "bad trades", especially since the only reason to request these in the first place is to guarantee your trade doesn't go through when abusing the dupe bug.  The server could simply drop the user's connection in this case, forcing them to reconnect if they want to make a "real" trade.  ''Note: This is the one fix I think <u>can</u> be implemented right now''
* Server should ensure that the previous transaction completed successfully on the client side.  The client can't be trusted, so double-check on every new transaction that the last one was completed on the client side before accepting another.
Now, my thoughts on making the search better.  These assume that the "impossible" trades have been blocked on the server side
* The biggest aide to search would be to provide results only for trades the player can fulfill.  The client could send a simple list of what pokemon the player possesses.  There's what, less than 512 pokemon currently?  That's 64 bytes we'd have to send with every search, definitely not a bandwidth-intensive process.  Server-side would filter the trades to only ones requesting a pokemon the player has.  All results are sent back, and the client side filters out ones with levels the player doesn't hold.
* 7 results?  WTF?  How useless would Google be if it only returned at most 7 results?  That's a load of crap.
* Cache the results on the client to speed up searching for more specific conditions after a generic search.
* Searching needs better level filters.  "10 and up" works fine for requests, but if the player wants a lv15ish pokemon they have no way to find only 10-19 levels... If I want a 15 do I care about the lv90 that's up there?  I doubt it.
Unfortunatly, I don't believe the D/P games can be patched from the server.  I assume that if the card had extra EEPROM to store patches in, many of these issues would have been fixed by now.  Bare minimum, I'd expect ''something'' to be done about the dupe bug.  That would be a super high priority, user interface be damned!  I hope that the designers learn from this pair of games and implement a much better GTS in the next games... and add some extra flash memory to those cards specifically designed for patching bugs that crop up after release.  [[User:Tekkub|Tekkub]] 20:03, 18 June 2007 (UTC)
1

edit