Concern about Randomness of Shuffles

[Greg the Yeti]

Go go go Richard!
why not share the tool with everybody?
Its so cool - tool!

Regards Greg

(Jason Statham again!)

[Greg the Yeti]

Hi Richard,
just went to start the cooking and thought:-

If having the mouse steady can provoke a non random result - that boosted #72, then how to guarantee randomness for absolute nuts like me??

What - Kill us??

Nice one! - That would generate no complaints and put us out of our misery - - - Very good!

But seriously. In test 2 that I did above I deliberately chose to do some activity that I reckoned would be pretty random. Though married to the quest for random-icity, it would be divorced from any prospect of being connected (married).

How to guarantee that every time? I was thinking that if the program suggested, for those of us who want genuine random draws, that you "Go away and play a little computer game for a few minutes while it collects its mouse movements". I was thinking specifically, something on the lines of "Bubble Breaker" - have you seen it? It's very simple, engaging for a short slot of time. Who knows you might even get a high score!??

It might even prove useful (at some stage in the late 25th century) to have such a game built into Orhpalese, specifically to encourage people to get their random Card Draws. Then we could even share high scores here on the Forum!

A zaney idea, I know, but it would work...

Regards Greg

[Programmer]

To clarify the simulation button, that is only relevant when you have selected shuffledebug.xml in the dropdown. It uses the exact same method used in the normal shuffle in the program, you can just start the simulation and it will run continuously until you press stop, you could leave it on overnight if you wanted to.

To reset and start again with either set of tests, just delete xml file. The program will generated a new one if it isn't there.

Holding the mouse steady to deliberately produce a non random result should not work because the program ignores mouse moves that are less than a certain minimum distance.

Building a game into the program? Hmmm.... I get the idea though. I think it should be possible to give the user the option to capture all the mouse events at windows level, even if the program was minimized or minimized to the system tray. Then the user could build up usable random data when they were filing their tax returns or playing some first person shoot 'em up or whatever. And then maybe be able to always opt to use that kind of data instead of the pseudo-random stuff if they preferred it. It is another one for the maybe list.

[Greg the Yeti]

Hi Richard,
thanks for your reply.

I had just been playing with the Random tests and actually tried deleting the ShuffleDebug.xml - hoping that it would do just as you say - generate a new one. It didn't do it, but generated a "try to continue or quit" type error!
My second attempt to "Force a Restart" was to go into the xml file with TextPad and edit it with the replace tool.
I selected all examples of <Counter>...</Counter>
and replaced with <Counter>0</Counter>
This did generate a file which looked identical to what I would expect a Fresh file to look like.
Unfortunately that also generated a program error. "Could not do something to do with "a0"???

(A little possibly relevant detail. As I was working in a seperate folder "Hide Me", with the original file right there I renamed the file ShuffleDebug.edit - but TextPad auto added a little to that so it came out as ShuffleDebug.edit.txt - with the regular warning about changing file extensions!
I don't think this should have changed the content of the file in any way? You might know about that?
Obviously I changed the file extension back to xml in order to test it.)

I replaced the original file (from the "Hide Me" folder I had put it in) and still got an error "Can not do something with "a0".

I shut down Orphalese and restarted and got no errors this time. ShuffleDebug.xml is continuing with its grand test (for all of time! - joke!)

However on just going back to check something, the error happened again. Clearly there is something here that it doesn't like!
- - -

I also note on the positive side however that "MouseCapDebug.xml" was working and accumulating data after just having started the Program. It looks to me as if this is how it happens and earlier I had not seen movement in the numbers because it already had enough random data and wasn't "accumulating"?

Can I suggest that it might be very nice if the program could be "Up Front" and indicate in some way when it is "accumulating" and when not.
Perhaps a little Orange or Green button in the Title Bar?
Orange - still accumulating
Green - All done for now
With a little pop up text to explain that to interested users?
(You might have done it already and I havn't seen it yet....)

Hope you like that idea...
Regards Greg

Perhaps:-
If I re-install Orphalese and get that totally "Fresh File" that way and then put it in my "Hide Me" folder, before even running Orphalese for the first time??? Its a bit over the top but might work?
Or Perhaps:-
Could Orphalese be "Parsing" ALL the folders within "Orphalese Tarot" and ALL xml files it finds - and producing the error that way?

[Programmer]

When the program writes to those debug files it is creating a file lock. I didn't explicitly say it but if you wanted to reset the debugging by deleting one of them the procedure would be:

1) Close the program.
2) Delete the file in question.
3) Restart the program (the program will recreate the file automatically when needed).

The tool is just a one off device to lay to rest the suspicion that the shuffle might be producing non random looking results. There are no plans at present to make it a regular part of the program, so obviously at the moment there is no plan to make any enhancements to it.

As I mentioned on a different thread I have now come to the end of the programming cycle for this version, but please do keep bug reports coming if you happen to see anything not working as it should.

Cheers!