Pallet Placement

[ArtF]

Hi Mark:

I did slay that dragon, did it come back to roost? I did a search after the repair and found the proper sequence you were missing,
did it screw up on the same search or another?

Art

[marklazarz]

First of all, Merry Christmas!

Regarding the dragon coming back to roost, something weird is going on.  I checked the About splash screen and it said v2.30050 so I figured I messed up the installation.  I deleted the shortcuts and any previous installation downloads and re-downloaded the current version.  After installation over the old version, it put new shortcuts on my desktop.  I ran the new software, checked the splash screen and it reported that it was still v2.30050.  During the installation I received a dialog box asking if I wanted to install v2.3006a (or something like that) and I answered yes and finished the installation.  The last part of the installation asked something about 32 bit runtime and I answered yes to that also.

I'm running Win10, 64 bit, Intel I5 processor.  I allow unsigned software installations and disable my antivirus software prior to installing Gearotic.  Should I save my user files and delete the Gearotic Motion directory and start over?

Thank you for your excellent, prompt and courteous replies to my questions!

Mark

[ArtF]

Hi Mark:

  No, totally my fault. I didnt update the version number, I have no doubt your on the correct version, Ive updated for next
update. When you see a letter after the version inthe program ignore it, it just means I did a minor update which may not be reflected in the program itself. ( Files etc..).

  Do a check for 1/60 , does the proper sequence you were looking for ( I forget it top of head ? 8 8 / 62 64? ) come up?

Art

[Mooselake]

Art, I don't know what you're using for source code management, but in the distant past when I worried about these things they would automatically update strings for version, date, etc.

Kirk

[marklazarz]

Hi Art:

I do not get any 2 gear - 8 8 combinations using autotrain with the selectors set to 8 min, 96 max.  I tried changing from involute to epicylcoidal, tried changing the mod, tried inverse ratio, tried various error limits, all with no luck.  Changing to min 8, max 64, no solutions are found.  My original question asked about 8 8 / 60 64.

[ArtF]

Mark:
Yes, dragons came back.. odd,I ll do some searching..

Art

[ArtF]

Kirk:

  lol, you give me far too much credit. Im a very old school programmer, I use no version control, no checkin - checkout system, no management at all. I poorly document, break all programming rules, and any teacher would likely take me to the woodshed. A result of programming for way too long, being self-taught, and just being who I am... Im prolific, but terrible at following rules. Problems such as this one always cause me to have to relearn how I did whatever algorithm it is I'm troubleshooting.
  While this causes me great frustration at times,I never learn... to thine own self be true... :-)

  This problem turned out to be one of rounding, varied a bit from system to system too, its now fixed ( or seems to work well and show many more options in the calc. ). This really was caused by me trying to be clever, the original algorithm someone sent to me to do the train calculations of ratios tool several minutes per search as there can be billions of possibilities for each search, I was kinda proud at mine taking only seconds or fractions of a second. No wonder, I was missing many entries at an error of 0. Im now getting twice as many entries on that search for example, though the time did double...Still, its very fast so Im happy with the result.

( My thanks to the originator of the Gear Calculations too, without him there wouldn't be one, he was kind enough to send me a search algorithm example for me to build on. I have a number of letters on the popularity of that calculator.  )

  Fix will be in next release update, likely within days.

Thanks Mark, for the double check...

Art

 

[marklazarz]

I'm not a programmer like you guys but I found it interesting that a 3 gear train set offered the sought after solution with an 8 tooth to 8 tooth middle set, effectively delivering the same ratio I was looking for but with a reversal of direction.  Perhaps you could use that information with your fast algorithm testing for same middle sets and push it over to the two gear solution table?

[ArtF]

Hi Mark:

Good thought, but the 3 solution model worked because it juggled 3 index databases, it found itself
deeper in the loops, so it found all the answers, the 2 gear model figured it found what it needed, so
it didnt look deep enough.. I think Ive beaten it into submission.. bruises on both sides..

Art