New Software. FTP HUD with 3-bets %!! Free!!!

[_dave_]

Mike,

This looks really good [img]/images/graemlins/smile.gif[/img]

[ QUOTE ]

Also it doesn't connect to the internets and tell me what your hole cards are


[/ QUOTE ]

Nice to know, however

[ QUOTE ]

If anybody who knows about this stuff could verify that


[/ QUOTE ]

Is way beyond my skills without source and compiler/settings [img]/images/graemlins/smile.gif[/img]

I certainly don't mean to imply anything bad by that, just that theres no way I could verify such... maybe Juk with his mad ASM skills could [img]/images/graemlins/smile.gif[/img]


A few thoughts...

No river stats? decided no value to them? I like to know who to value bet hard / fire 3 barrels at, although I suppose this does take large sample size.

Is there some problem with the screen shot? or am I blind... I only see the detailed pop-up.


Again, looks awesome - will prob check it out in the next couple days [img]/images/graemlins/smile.gif[/img]

Many thanks,

dave.

[mikechops]

Yeah I wouldn't know how to verify non-holecard transmitting either. I think you could probably do it by seeing whether Realtime opened up any sockets. If it doesn't then it isn't talking to anybody. The nitty-gritty of how you do this is beyond me though.

Alternatively you could just trust me. I've been around here a while [img]/images/graemlins/smile.gif[/img]

I could add specific river stats, but I think they are less useful. For one thing you don't get as many samples, but mainly because the way the hand has been played out up to that point influences the river decisions enormously. For now you just have the general agg, bet/call, raise/call/fold and c/r stats for the river. But if there's a lot of demand for them...

Yeah sorry the screen grabber didn't show the player panels. Imagine gametime+ with colors and an option to show stat abreviations.

[demon102]

anyway to get this working for stars? I have no rb at ftp

[mikechops]

Stars is next on my todo list. It don't think it will be take long, but will probably be fairly tedious. I'm going away next week and might get to it when I get back in early July.

[_dave_]

[ QUOTE ]

The sad thing is that it's almost impossible to be 100% sure for any reasonable size application. In theory it's possible to pack something away deep inside a segment and then have some code that unpacks and/or drops it based on a semi-random condition being met.


[/ QUOTE ]

Thus the need for source + exact compiler version, flags etc - if one can read, understand the source thoroughly (probably can't, but after many reads I kinda get FPHG 0.x) - then one can compile in identical environment which should produce an executable with matching MD5, SHA or whatever sums... Note I probably don't know what I'm on about much here [img]/images/graemlins/smile.gif[/img]

Of course it may not be a worry at all - If this does no injecting / hooking and simply parses HHF files, it can be effectively quarantined as you say with ProcessGuard, snoopfree, and a app level firewall.

My concern (Not really a concern given this is MikeChops, posting under real account - multiply paranoia by * 1000 if new reged user on 1st posting [img]/images/graemlins/smile.gif[/img] ) Is similar to your not wanting to release FPHG source - if it injects to FTP for whatever reason, theres almost no way to tell what it be doing, without source comprehension.


Mike, I will try this, and it looks most impressive. Do not think I am doubting your generous contribution, I certainly am not.

I just wish it was open source [img]/images/graemlins/smile.gif[/img] How times have changed post RBCalc.exe [img]/images/graemlins/frown.gif[/img]

dave.

[_dave_]

Mike,

On a less paranoid note [img]/images/graemlins/smile.gif[/img]

How are you storing the informations? custom DB? 300K HHF in 5 minutes (IIRC from the readme) is impressive... but surely you are storing for future start ups.

So, standard DB? or just summary files? As in, does this produce another data-source we may write custom queries for?

dave.

[mikechops]

Yes a file is created and stored for each game type (Holdem NL, Holdem FL, Omaha PL...etc) It is a custom format, but if anybody wants to know what's there, I'll send them the Load/Save code.

I scan the HHFs and keep a record for each player of everything that's in the stats window and each game #. So the format could change if I add additional stats. I guess it might be better if I had saved the length of each player entry at the start of the DB. That way I could add additional stats onto the end and if anybody wanted to read the data, they could just ignore it stuff that was added after they wrote their prog...

There's also a Save As Text menu option. It dumps the stats for each player as one huge text file.

[the pang]

Did you ever finish realtime for Poker Stars?

[VP$IP]

[ QUOTE ]
Did you ever finish realtime for Poker Stars?

[/ QUOTE ]

I don't think it will ever be "finished", but we are all using it!

[jukofyork]

[ QUOTE ]
I certainly don't mean to imply anything bad by that, just that theres no way I could verify such... maybe Juk with his mad ASM skills could [img]/images/graemlins/smile.gif[/img]

[/ QUOTE ]
The sad thing is that it's almost impossible to be 100% sure for any reasonable size application. In theory it's possible to pack something away deep inside a segment and then have some code that unpacks and/or drops it based on a semi-random condition being met. It could take weeks/months/years on a decent sized executable just checking execution paths to make sure this doesn't happen (not that I think their is anything wrong with this app or that many would go to this much trouble to hide a Trojan...).

Blocking stuff from Internet access and using something like ProcessGuard (or a DLL injection blocking firewall) is probably your best line of defence, although some virus scanner may be able to pick up on potential new threads by using a heuristic scoring system (this tends to work better for new [or polymorphic] viruses though, as they have a more well defined set of functional blocks that must be implemented for each and every virus).

Juk [img]/images/graemlins/smile.gif[/img]