unknown balances cause delay or no action. Backwards Incompatibility

[GoogleCodeExporter]

OpenHoldem PRB2

Problem occurs when:
1. Playing at a casino where the players last action Call/Bet/Raise covers
their name and balance temporarily. 
2. TM that cant/doesnt scrape one players balance.

OH will not act until it can see the balance/name. This is different than
OH 1.x where the player would act.

The relevant code is CGameState.cpp ::ProcessThisFrame:
<pre>
   // check if all balances are known (indicates stability of info passed
to DLL)
   balance_stability = true;
   for (i=0; i<10; i++)
   {
      if (_m_holdem_state[(_m_ndx)&0xff].m_player[i].m_cards[0] != 0 &&
         _m_holdem_state[(_m_ndx)&0xff].m_player[i].m_cards[1] != 0 &&
         _m_holdem_state[(_m_ndx)&0xff].m_player[i].m_balance_known != 1)
      {
===>      //balance_stability = false;
      }
   }
</pre>

Possble suggested solutions:
- option in TM file
- remove entirely
- option in .ohf file

I prefer the option add to .ohf file. My arguement would be that whether or
not you need the balances stable is a bot wide option. If your bot needs it
at one casino it needs it at all. If put in TM's, all tm's would probably
have the same value.

Links to where testers have possibly experienced this issue:
http://www.maxinmontreal.com/forums/viewtopic.php?f=120&t=7474
http://www.maxinmontreal.com/forums/viewtopic.php?f=110&t=7383
http://www.maxinmontreal.com/forums/viewtopic.php?p=52270#p52270

Original issue reported on code.google.com by Oskar.Mo...@googlemail.com on 5 May 2009 at 2:27

[GoogleCodeExporter]

Tagging as enhancement.  We still believe that the root cause of this should be 
fixed
- in other words, setup your tm such that invalid balances do not occur.  OH 
will
recognize "All-In" as zero balance, if scraped properly.  Hashing a "blank" 
balance
box and returning zero will handle those situations.

Regardless, an option will be added to preferences to enable this feature - 
default
will be disabled, as per 1.x behavior.

Original comment by openhol...@gmail.com on 13 May 2009 at 3:40

• Changed state: Accepted
• Added labels: 2.0.0-PRB2, Type-Enhancement
• Removed labels: Type-Defect

[GoogleCodeExporter]

Preference added, rev. 804.

Again, we strongly recommend that you address the root cause of the problem 
with the
Table Map, instead of relying on this preference as a kludge.

Original comment by openhol...@gmail.com on 13 May 2009 at 4:34

• Changed state: Verified