#21
|
|||
|
|||
Re: How do I speed up PT importing?
[ QUOTE ]
[ QUOTE ] [ QUOTE ] [ QUOTE ] PT3 and HM are being released soon. On PT I average 10 hands/sec and on HM I'm doing 50 h/s. [/ QUOTE ] That's still slow, PokerAcademy Prospector (using sqlite BTW) has recently set the benchmark at 300 odd hands/sec. Josh and Roy have some catching up to do. [/ QUOTE ] Just to clarify Poker Academy Prospector doesn't use sqlite, but a custom solution. [/ QUOTE ] Custom solution? That means there will be no third party add-on applications? I see that as a major problem. [/ QUOTE ] If there is an API, it doesn't matter what the underlying mechanism is. There is no API yet, but we really hope to make things easy for third party applications soon. |
#22
|
|||
|
|||
Re: How do I speed up PT importing?
[ QUOTE ]
[ QUOTE ] [ QUOTE ] [ QUOTE ] PT3 and HM are being released soon. On PT I average 10 hands/sec and on HM I'm doing 50 h/s. [/ QUOTE ] That's still slow, PokerAcademy Prospector (using sqlite BTW) has recently set the benchmark at 300 odd hands/sec. Josh and Roy have some catching up to do. [/ QUOTE ] Just to clarify Poker Academy Prospector doesn't use sqlite, but a custom solution. [/ QUOTE ] pah! My Realtime HUD imports 250k hands in under 250 secs. Of course it does help that most of the hands are 6-max NL. If I had to import a lot of 9-max FLO8, it would probably be a lot worse. [/ QUOTE ] Very impressive! We do a lot of error checking in Prospector though too and fixing bad hand histories so the ones that make it through the import are all valid which hurts the import speed a lot. Taking that code out makes it significantly faster and much more like what you are seeing. Have you released any details on your HUD? From the bits and pieces you have mentioned here and there it sounds very promising. |
#23
|
|||
|
|||
Re: How do I speed up PT importing?
[ QUOTE ]
[ QUOTE ] [ QUOTE ] Just to clarify Poker Academy Prospector doesn't use sqlite, but a custom solution. [/ QUOTE ] You sure about that? the sqlite dll in my Prospector\data directory disagrees with you. [/ QUOTE ] That's for PokerRoom importing. [/ QUOTE ] Penny drops. Terence, I knew I recognized that surname. Thanks for clearing that up, saved me some time trying to untar the xzf. |
#24
|
|||
|
|||
Re: How do I speed up PT importing?
[ QUOTE ]
[ QUOTE ] [ QUOTE ] [ QUOTE ] [ QUOTE ] PT3 and HM are being released soon. On PT I average 10 hands/sec and on HM I'm doing 50 h/s. [/ QUOTE ] That's still slow, PokerAcademy Prospector (using sqlite BTW) has recently set the benchmark at 300 odd hands/sec. Josh and Roy have some catching up to do. [/ QUOTE ] Just to clarify Poker Academy Prospector doesn't use sqlite, but a custom solution. [/ QUOTE ] Custom solution? That means there will be no third party add-on applications? I see that as a major problem. [/ QUOTE ] If there is an API, it doesn't matter what the underlying mechanism is. There is no API yet, but we really hope to make things easy for third party applications soon. [/ QUOTE ] I'd still be worried about a custom solution. These databases have been thoroughly tested and the chance of data loss is extremely low. Not to mention they offer more configuration options, such as having the server run on a different computer than the poker tracking software. Also, through ODBC, any programming language can access the database. With a custom solution, they would have to hack around with a C interface, or you'd have to provide interfaces to a lot of different languages (or your own ODBC driver). The only time an ultra fast import speed really comes into play is with the initial import. After that, importing while you play only does a few hands per minute. Writing a custom solution just to gain this tiny feature point, isn't worth the negatives, IMO. |
#25
|
|||
|
|||
Re: How do I speed up PT importing?
[ QUOTE ]
The only time an ultra fast import speed really comes into play is with the initial import. After that, importing while you play only does a few hands per minute. Writing a custom solution just to gain this tiny feature point, isn't worth the negatives, IMO. [/ QUOTE ] Agreed. |
|
|