MarketFeeder Pro 8.0 - BETA testing

Hi

Np. See the attached.



Larp

Hi Oxa,

I had several problems trying to get 8.0.0.7 working reliably this morning.  It took several attempts to get the program running and remaining running.  When I started another instance using a different profile one of the instances would crash.  After several attempts I managed to get two instances up and running (on Windows and they have run all day (~12 hours), one in test mode, one in real mode.  So the start-up seems a bit flaky.

Now a plea for a small change if possible.  A low priority given some of the other feedback I'm sure!

Below are 2 examples taken from a Test Statement and a Real statement (as created today from the 2 instances).  Is it possible to make the format of the test and real statements the same please?  I load the statements into Excel to do post analysis on the bets to refine and check my systems.  Now I'm moving from test to real mode for some of my systems, I now have to combine the statements in Excel and the fact the formats are different is a pain because I'm extracting track, distance and race type from these strings.

TEST   21:21:59   Horse Racing / Wolverhampton / 21:15 Wolv 15th Nov 1m Hcap - 1m Hcap
REAL   17:48:24   Horse Racing / GB / Wolv 15th Nov/ 17:45 5f Hcap

I'll put an IF statement in my Excel spreadsheet meanwhile to test if the string is real or test.

BTW, the issue I  raised previously regarding multiple bets not being placed by the maximum (of all) condition would appear to not be an issue after all.  Multiple bets are being placed so your comment about the odds moving must have been the cause.

regards

Wez

2nd attempt at posting....

I found the attached trigger solved the problem of updating test statements once the market had settled.  It serves no purpose other than to cause itself to be executed every 30 secs.  It worked for me and I've used it ever since every time I create new trigger files.

Just posting in case it helps other with a similar problem.

regards

Wez

I found the attached trigger solved the problem of updating test statements once the market had settled.  It serves no purpose other than to cause itself to be executed every 30 secs.  It worked for me and I've used it ever since every time I create new trigger files.

Thanks for the tip about getting test statements to settle Wez, I'll try that today.  UPDATE: I'm still getting lots of markets unsettled despite adding this trigger to my test trigger, but I'm glad it's working for you.

I had several problems trying to get 8.0.0.7 working reliably this morning.  It took several attempts to get the program running and remaining running.  When I started another instance using a different profile one of the instances would crash.  

Unfortunately I had this happen this morning with version 8.0.0.7.  It hasnt happened with any earlier release of version 8.  The behaviour was similar to what was happening towards the end of version 7 where something was conflicting with microsoft updates to internet explorer.

Below are 2 examples taken from a Test Statement and a Real statement (as created today from the 2 instances).  Is it possible to make the format of the test and real statements the same please?

While we're talking about statements, if you are going to tweak them Oxa could I possible ask if my long running request to reinstate the date in the time column could be actioned.  As a reminder, this is how a statement looked in version 6:

03/01/2012 15:35:40   BACK   Horse Racing - Todays Card / 15:30 Sthl - 6f Hcap

By adding the date into the time column it allows for much easier record keeping, as it's far easier to sort results by date to see how triggers performed over time.

As Wez said, I realize statement tweaks are low priority at the moment.

I blamed myself above for the transaction charge, but thinking about it, I am running all the same triggers and platforms as I did prior to the API change, mainly V6 (good and reliable) and some V7.  Now all V8 of course.

I had my first problem starting V8 yesterday getting an error about a "logging port" being already in use and to change the port number in the settings - although I have no idea how to do that - but after ignoring the message it all started up ok but I was missing some of the log.

On Friday when  the data charges we for I had had no error messages and I believed all were running fine.

( thanks for sharing the settings)

Thanks Oxa for the market_country variable.
It's reduced the amount of text that gets logged compared with when i used to use IF(FIND("GB","market_name")>0 etc

Im having an issue with bets placed on US races being cancelled at Inplay rather than taking SP, anyone else having this?
Using all same settings as previous with MFP7

Hi

I have a virtual private server with three different instances of MFP8 (v 8.0.0.7). I recive frequent error messages "feeder8: abnormal program termination", or the program quit without messages. I thought it might depend on the large number of markets and then I down the time interval in market locator, and run the market locator every hour, but the error persists.

I never encountered this problem with MFP7.

Thank you for support
Diego

I noticed that an unexpected shutdown occurs frequently in AUS & NZL greyhounds  races

I have a virtual private server with three different instances of MFP8 (v 8.0.0.7). I recive frequent error messages "feeder8: abnormal program termination", or the program quit without messages.

Hi Diego,

If this post is related to another one of yours:

I noticed that an unexpected shutdown occurs frequently in AUS & NZL greyhounds  races

then we've found the API change that accounts for this behaviour. Obviously the new API reports some runners as "Losers" even before the start of the race - no explanation for that, but this is unexpected and makes the program crash. I am working on a fix, please watch this topic for an update.

While we're talking about statements, if you are going to tweak them Oxa could I possible ask if my long running request to reinstate the date in the time column could be actioned.

armarni, but isn't the date already stated in the top-right corner of the statement page, as each date has its own page appointed?

armarni, but isn't the date already stated in the top-right corner of the statement page, as each date has its own page appointed?

I, (and I think a lot of people) trial different strategies by using different bet amounts so they can keep track of which test strategies are working and which aren't.  This means every day you have lots of results which need entering into a spreadsheet eg all the bets that where placed at £2.01, all the bets that where placed at £2.02 etc.

If you do this each day religiously then that's fine, but if you have a few days or weeks worth of results to enter at once then without the date in the time column, you have no way of sorting all the results by date, as in the statements the date info is buried in each races name and any sorting of that column sorts by course name, not date.  

As an example, lets say you have 10 days worth of results to put into the spreadsheet.  Without a date column each days results would have to be entered individually to keep them in chronological order, which if you're testing 40 strategies means entering 40 results per day,  eg all the £2.01 bets from day 1, all the £2.01 bets from day 2 etc.  However, if you can sort the different bet amounts by date, then you can merge all the csv statement files and enter all the results for each bet amount at once eg all the £2.01 bets from the 10 days in date order, all the £2.02 bets from the 10 days in date order etc, speeding up the data input immensely.

How I proceed at the moment is I add a column to each days statement with the date in, so each entry has a date cell in front of it.  I then merge all the statement csv files that need entering into my results spreadsheet and can then sort the individual bet strategies in date order (eg all the £2.01 bets in date order, all the £2.02 bets in date order etc).  

Adding the date column manually is what I ve been doing for the last few years, but when you have 30 days worth of previous results to add to the spreadsheet it's labourious and adds a lot of extra work to the process, and it's frustrating as the date used to be in the time column in version 6, so I was just hoping it could be reinstated.

Hi Oxa,

When do changes to the settings get recognised?  I booted MPF this morning, changed the default display setting to Engineer mode with some default values and the changes still haven't been applied even though I have rebooted the program.  Also, after the first boot, I changed the view to engineer mode for one race (becasue the settings weren't doing it...) and whilst I could edit an engineer mode field, when I closed the dialogue box the change hadn't taken affect.  If I re-edit the same box, it still says 0 even though I had just changed it to silk_weight!

Also, minor typo in help file regarding "silks".  It says "Returns 1 if silks are supported in this market, or 1 otherwise."

Should say ".... 0 otherwise" presumably!

And on a pedantic note  .  In the Excel tab in Settings it says "Warning: You do not need to have MS Excel installed....."

It's not really a warning is it?  Isn't it "Note: You do not need to have MS Excel installed....."  Not needing MS Excel for MFP to operate is a positive isn't it?!  

Regards

Wez

Today I launched in Betfair Mode, with the same profile but yesterday's test statement and previous test statements using MFPro8 were not available.  From this evidence it appears that after updating the program, test statements are lost.  

Oh no, you can't share the same profile between BetFair and Time Machine. Or rather you can, but it makes no sense, strongly discouraged. Of course your statements would be missing after a round of bets in Time Machine. Your markets would not pick up either.

Hi Oxa

I get this in my clean installation of V8, in my clean vps (no any Feeder versions installed).

It is needed??

Thanks
Best regards
pcal72

pcal72, no, it is not needed, I will see how it got into the installation.