Yep, those two "squares" have always been their though - I just checked some really old HH I mined a few years ago and they had it too so it's not something that's just been added/changed recently.

I'm not sure why HM isn't importing the hands, but I don't think the squares are to blame.

Juk

I reinstalled Party Poker Software and it seems to work now.

Thanks for your help.

Dunno if this is significant at all, but after the January party update it was crashing circa once in every 500 hands. Then I moved my fphg output folder to a RAM drive, and instructed HEM to auto-import from there. After that change, party has crashed only 3 times in 10k hands. Could be coincidential though. Anyway, dunno really how fphg works exactly but could the crashing/freezing have something to do with fphg being unable to write into the live hh because HEM/betpot ahk is currently reading it? And moving the hh folder into RAM made the "collisions" less likely, because reads and writes are much faster.

Yep, I thought this might have something to do with it (the times I tested it then I not had anything reading the output files like HM, etc) but then Peter tested it without anything reading the HH files and said it still crashed. The best way to test if it is this would be to just to run FPHG without anything reading it's output for a session or two and see if that works. If this is the reason then it wouldn't be too hard to fix.

Another reason why a RAM drive might also alleviate things is if the disks writes are causing the process to stall somehow. If I get some time and nobody has got to the bottom of this I'm gonna rewrite the whole of FPHG v2 so as to make the bare minimum amount of work get done in Party's EXE by passing the data out asynchronously to another application which will do all the hard work. This should mean that the minimum extra work is done inside of Party's EXE at the cost of a very slight delay between the Party game state and the writing of hands.

Juk

Just wanted to add an update regarding RAM disk usage. I guess it was only pure luck that at first the crashes seemed to happen more rarely, but now after 2 weeks I can safely say that using RAM disk doesn't help at all.

What is strange though, I can observe 18 tables while HEM is importing and HUD is on, and it doesn't crash. Only when I'm playing.

Did you try running without HEM reading the ".live" files? If so, did that make any difference?

Juk

Does HEM read the .live files if I have added the fphg output directory to HEM's auto import folders? I was under the impression that HEM can read the .live files only from the fphg's default directory, and I have changed that.

I don't have HM so can't really say for sure, but it might read the FPHG.ini file or something and track down where you are saving the hands (perhaps ask HM how they are doing it).

Juk

Hopefully this is addressed elsewhere but FreePHG seems to be picking up my played hands (and not only observed hands) as intended. The ini is set up to grab only observed hands:

OBSERVED_ONLY = 1

It's causing all kinds of problems with PokerTracker - truncated half-files and so on. What am I doing wrong? Thanks.

If you open the hand history files that FPHG is writing do they stop after the "Dealt to XXXXXXXXX [ XX XX ]" line? If so then PT should ignore them (even if it prints a few import errors) and it's always worked like this (a bit of a hack as I you can't tell if it's an observed hand or not until you see that line). I've never tested this on anything other than PT2 though, so it could potentially cause problems in PT3 and HEM.

If on the other hand FPHG is just ignoring the option then try manually deleting all of FPHG's folder and reinstall it (sometimes old versions can get the DLL locked, etc and end up with a different DLL to what it should have).

Juk

Nope, they look like this:

Yep, they are truncating as soon at the "Dealt to " line is seen - this is normal, although a bit hacky. Are you importing these into PT3 and it's causing problems? It never effected PT2, but could well effect PT3.

I am in the process of rewriting FPHG atm anyway (to try and fix the phantom crash bug that some people get) and adding a flag for realtime output is something that's always been planned (and will fix your problem).

In the mean time you can try using the old version of FPHG: http://www.jukofyork.com/FreePHG_v0_07c.zip

If you edit the batch file to use the "-o" option then it will output only observed hands without any truncated ones (beware though the method used to grab the hands can be slow and more CPU intensive compared to FPHG v2).

Juk

Quit FPHG (it was hung, btw - force quit the process), reinstalled it, reset the .ini settings, ran it again, sat down at a table. PT output looks like this:

As you can see, PT3 is picking up files from both locations FPHG seems to be ignoring the observed only configuration.

A workaround is to turn off the hand histories from inside party, and just let FPHG do all the work, but this is suboptimal: if there *are* any differences in the Party hand histories (even at some point in the future), I don't want to have to remember to download the latest FPHG.

Great software, btw.

Ah, because it doesn't need to look for my involvement in a hand otherwise. I'll just turn off the hand histories from inside Party and let FPHG do all the work. I look forward to a new version, though.

(total aside: your web page background has *got* to go. my eyes, they burn)

This would work fine as all FPHG does is grab the hands from Party's memory space (it doesn't format/create them itself) so there will never be any difference between the FPHG outputted HH files and the ones Party output's itself.

Juk

Hehe, my whole website is a horrible nightmare. One day I'll sort it out, but atm it's just this gigantic mess of random HTML files, RARs and god knows what else all edited by dropping files into the root folder via FTP....

My brother was supposed to be finding me some nice PHP based content manager for it, but never got round to it. I'll pester him again if I can.

Juk

Hey Juk,

I'm having an issue with FreePHG v2.14 and the latest party update. The issue is that some HH that are put on the HDD have errors in them. Here's two examples:
As you can see the river action misses a check and the showdown is wrong as well. As you can imagine this messes up the trackers and HUD (HEM in my case). This happens pretty frequently as well, I would guess one every 50 hands.

Any help will be greatly appreciated!

I'll have a look, but the best thing is probably to wait until I release the new FPHG as the bug could be related to the crash-causing one.

Juk

hey there juk,

re: the crashing
a few observations that hopefully might be of use.

when simply mining with fphg i have few crashes (when i don't try and look at the table being mined), however if i try to play while fphg is running it is a whole different story.

i have never made it more than half-way through a sng with fphg running and other tables open. usually within 5 min or so party will lock-up and go to 99% cpu resources (i kill it with tm and re-start party).

in addition, while running a mining script (e.g., safemine) direct mouse clicks on the tables being mined seem to cause the same problem. if i try and look at the tables being mined by maximizing them (and/or minimizing or closing them afterwards) without first pausing the script, then this causes the same lock-up scenario a high proportion of the time. whereas if i pause the mining script first, i get few lock-ups when minimizing/maximizing and opening/closing various tables.

cheers,
sean

ps. you sure look good in green!

Hey Juk,

Have you made any progress since then? Lots of hands are still incorrectly imported into HEM which is very annoying.

Can I help track down this issue in some way?

Thanks.

No sorry, I've been busy with other stuff lately and not had time to do any more yet.

Juk

hello,

i have a few hhf from fphg that seem to be a little messed up, thought they might be useful when you get the time to have a look at fphg.

these first 3 are examples of the garbled-type of fphg error, which often cause database errors in PT2 (i have log files if useful):

**** Hand History for Game 8099615138 *****
NL Texas Hold'em $11 USD Buy-in Trny: 45817691 Level: 6 Blinds(300/600) - Wednesday, May 27, 02:29:56 PDT 2009
Table Table 127799 (Real Money)
Seat 5 is the button
Total number of players : 2
Seat 5: juliak49 ( 3,420 )
Seat 9: vince080873 ( 16,580 )
Trny: 45817691 Level: 6
Blinds(300/600)
** Dealing down cards **
juliak49 is all-In [3,120]
vince080873 calls [2,820]
** Dealing Flop ** [ 9s, 2c, 6s ]
** Dealing Turn ** [ 7d ]
** Dealing River ** [ 5c ]
hows [ 6c, Kh ]a pair of Sixes.
of Sixes.
vince080873 shows [ Kd, 7h ]a pair of Sevens.
vince080873 wins 6,840 chips from the main pot with a pair of Sevens.
juliak49 finished in 2 place and won $30 USD.
eft the table.
ft the table.
Congratulations, vince080873! You won $50 USD.
vince080873 has left the table.
as finished, congratulations to the winners.
Table Closed
rs.
Table Closed

Game #8099641038 starts.

#Game No : 8099641038


***** Hand History for Game 8099657670 *****
NL Texas Hold'em $11 USD Buy-in Trny: 45818165 Level: 2 Blinds(30/60) - Wednesday, May 27, 03:03:39 PDT 2009
Table Table 127799 (Real Money)
Seat 6 is the button
Total number of players : 10
Seat 1: DarkLouis ( 2,080 )
Seat 4: Id1ot_Savant ( 1,600 )
Seat 8: LembitOpik ( 2,120 )
Seat 7: Milamber55 ( 1,860 )
Seat 3: ZehVini ( 1,980 )
Seat 2: arbndorga ( 2,560 )
Seat 5: brezzy01 ( 60 )
Seat 6: chaz2113 ( 2,500 )
Seat 10: confusedpoke ( 2,740 )
Seat 9: razvy8888 ( 2,500 )
Trny: 45818165 Level: 2
Blinds(30/60)
** Dealing down cards **
wn cards **
razvy8888:
razvy8888 raises [120]
confusedpoke folds
razvy8888: ya
DarkLouis folds
arbndorga folds
ZehVini folds
Id1ot_Savant folds
brezzy01 folds
chaz2113 folds
Milamber55 folds
LembitOpik folds
razvy8888 shows [ Kc, Ac ]high card Ace.
razvy8888 wins 210 chips from the main pot with high card, Ace.
Game #8099658701 starts.

#Game No : 8099658701


Game #8083878586 starts.

#Game No : 8083878586
***** Hand History for Game 8083878586 *****
NL Texas Hold'em $11 USD Buy-in Trny: 45738948 Level: 1 Blinds(20/40) - Wednesday, May 20, 23:10:20 PDT 2009
Table Table 126209 (Real Money)
Seat 2 is the button
Total number of players : 10
Seat 8: Ado_Bo ( 1,880 )
Seat 6: Danke_Man ( 1,480 )
Seat 5: Evil_Chukky ( 2,060 )
Seat 9: Fritten111 ( 1,820 )
Seat 10: Paula0712 ( 1,720 )
Seat 1: Randales ( 1,940 )
Seat 3: kascha59 ( 2,400 )
Seat 7: mtlhustler ( 3,180 )
Seat 2: nsb11622 ( 1,840 )
Seat 4: ziffle01 ( 1,680 )
Trny: 45738948 Level: 1
Blinds(20/40)
** Dealing down cards **
Evil_Chukky folds
Danke_Man calls [40]
mtlhustler calls [40]
Ado_Bo folds
Fritten111 folds
Paula0712 folds
Randales folds
nsb11622 calls [40]
kascha59 calls [20]
ziffle01 checks
** Dealing Flop ** [ 4d, 2c, 3d ]
kascha59 checks
ziffle01 checks
Danke_Man bets [100]
mtlhustler calls [100]
nsb11622 calls [100]
kascha59 folds
ziffle01 folds
** Dealing Turn ** [ 8d ]
Danke_Man checks
mtlhustler bets [240]
nsb11622 calls [240]
Danke_Man folds
** Dealing River ** [ Jc ]
mtlhustler bets [2,760]
nsb11622 is all-In [1,460]
cks.
nsb11622 shows [ Ad, 8h ]a pair of Eights.
mtlhustler wins 1,300 chips from the side pot 1 with a pair of Jacks.
mtlhustler wins 3,900 chips from the main pot with a pair of Jacks.
s from the main pot with a pair of Jacks.
nsb11622 finished in 10 place.




------------

these last 3 only report half of the hand (there were a bunch all together). these seem like they might be due to my computer lagging.



Game #8069614468 starts.

#Game No : 8069614468
***** Hand History for Game 8069614468 *****
NL Texas Hold'em $11 USD Buy-in Trny: 45667271 Level: 1 Blinds(20/40) - Friday, May 15, 13:51:19 PDT 2009
Table Table 131085 (Real Money)
Seat 10 is the button
Total number of players : 10
Seat 6: AnosmicAK ( 2,000 )
Seat 7: p0mmer ( 2,000 )
Seat 1: Stjerneskud3 ( 1,960 )
Seat 10: Stonefish71 ( 1,980 )
Seat 2: Terreur8888 ( 2,060 )
Seat 5: Tudazul ( 2,000 )
Seat 3: chouraciro ( 2,000 )
Seat 8: pimalin ( 2,000 )
Seat 9: robgoble ( 2,000 )
Seat 4: sinucao01 ( 2,000 )
Trny: 45667271 Level: 1
Blinds(20/40)
** Dealing down cards **
Game #8069616336 starts.

#Game No : 8069616336
***** Hand History for Game 8069616336 *****
NL Texas Hold'em $11 USD Buy-in Trny: 45667271 Level: 1 Blinds(20/40) - Friday, May 15, 13:51:56 PDT 2009
Table Table 131085 (Real Money)
Seat 1 is the button
Total number of players : 10
Seat 6: AnosmicAK ( 2,000 )
Seat 7: p0mmer ( 2,000 )
Seat 1: Stjerneskud3 ( 1,940 )
Seat 10: Stonefish71 ( 1,980 )
Seat 2: Terreur8888 ( 2,120 )
Seat 5: Tudazul ( 2,000 )
Seat 3: chouraciro ( 2,000 )
Seat 8: pimalin ( 2,000 )
Seat 9: robgoble ( 2,000 )
Seat 4: sinucao01 ( 1,960 )
Trny: 45667271 Level: 1
Blinds(20/40)
** Dealing down cards **
Game #8069620388 starts.

#Game No : 8069620388
***** Hand History for Game 8069620388 *****
NL Texas Hold'em $11 USD Buy-in Trny: 45667271 Level: 1 Blinds(20/40) - Friday, May 15, 13:53:15 PDT 2009
Table Table 131085 (Real Money)
Seat 2 is the button
Total number of players : 10
Seat 6: AnosmicAK ( 2,000 )
Seat 7: p0mmer ( 2,000 )
Seat 1: Stjerneskud3 ( 1,940 )
Seat 10: Stonefish71 ( 1,940 )
Seat 2: Terreur8888 ( 2,080 )
Seat 5: Tudazul ( 2,000 )
Seat 3: chouraciro ( 2,120 )
Seat 8: pimalin ( 2,000 )
Seat 9: robgoble ( 2,000 )
Seat 4: sinucao01 ( 1,920 )
Trny: 45667271 Level: 1
Blinds(20/40)
** Dealing down cards **
Game #8069624987 starts.



i can't offer any help as i am no programmer; however, if any hhf, log files or other info would be of assistance please let me know.

cheers,
sean

This could well be related to the crash bug. FPHG has always picked up on some dodgy hands though - even the original FPHG did this. The only surefire way to be sure the hands are OK is to parse them (atm it just does some very basic checks to see if it looks remotely like a HH file).

Juk

FYI, the issue that I'm having looks the same as theRealMacoy describes.

Is FPHG still being supported please? I'm getting a whole lot of "Invalid pot size.." errors at Party now.