DSP Robotics Support

The 2 "Menu's" i have in FS both dont update. . . ?


EDIT:

Selector updates. . . .

ahhh ... check to see if the output pins give correct values.

I was a bit surprised finding 2 similar items like this ... and yes, the 'Selector' version works how I kind of expected.

The other [Menu] one stills says Menu at all times, but the output values are correct ... crazy

What this means is that we now have 3 pull-downs to verify.

RJHollins wrote:ahhh ... check to see if the output pins give correct values.

I was a bit surprised finding 2 similar items like this ... and yes, the 'Selector' version works how I kind of expected.

The other [Menu] one stills says Menu at all times, but the output values are correct ... crazy

yes and this is correct! because when you have a menu like in the presetmanager you just want the button named
" menu " and apply one of the menu options i.e. save preset - means do something but after it is done you don`t want to see save preset on your button - you want to see menu again because the menu offers you some different actions,

while a selector is selecting an item or state which you want to see on the buttons surface... like you want to see wich filter is actually selected

if you connect the string output of the droplist with the label (inside the menu module) you see the actual item, and your menu is a selector now.

but be careful normally you want to save the state of a selector so you would need a preset option too while you dont need this in a menu

Thanks NuBeat for sharing the insight on these 2 objects. This now makes sense.

Trying to solve these issues .... well, at least I can say I'm learning [or being taught] some things


But now this bring to focus the Pull-Down menu test.

The Ruby module is a Selector. The PRIM is a Menu. Are we not comparing apples to mangos

RJHollins wrote:Thanks NuBeat for sharing the insight on these 2 objects. This now makes sense.

Trying to solve these issues .... well, at least I can say I'm learning [or being taught] some things


But now this bring to focus the Pull-Down menu test.

The Ruby module is a Selector. The PRIM is a Menu. Are we not comparing apples to mangos

yes you have a droplist prim from which you can build a "menu" or a "selector" i would say a menu provides you different triggers from a list and a selector let you choose a selection from a list

did you also tested the new ruby droplist?

something like this:
because i saw that the pdown.fsm from drnkhobo was just for a ruby label but not for the droplist methode

oops sorry ther was a little bug in the ruby droplist example, fixed it and reuploaded it

just did ....

When first loaded, had the Ruby error [red box].

Selecting any item cleared the error, related to a 'nil' condition.

Still inside FS ... obviously, testing on everyones system should happen.

Right now I'm putting another 'Text Save/Load' test together, as this function has failed on 2 separate systems [Win 7, and Win 8.1] for reasons unknown to me

For this test, I can only think of 2 different procedures. The use of all PRIMS, and then the Marshal RUBY [should try both standard and RB].

Since none of these have yet to work ... is there a 3rd way ??? Is there any way to add 'error checking & reporting' ??

I'm not really trained in this area ... heck, never had formal computer training [except for a taste in college writting code for a Z-80 processor to make it go BEEP Ahh ... Life in the '70s.

*** edit ***
re-download and testing

**** re - edit ---
Ya fixed it !

**** another EDIT

Since I confirmed that SONAR 8.5 had issue with the RUBY version ... maybe that was not really the case !!

The MENU just seemed frozen [not working] ... Now I can't recall if we also saw it's output during the test.

To complicate things even more ... my beta-tester had 'frozen' selector on my project. HOWEVER, I'm thinking this relates to the corrupt config file that feeds all these menus.

In light of all this, I would like to retract my SONAR trouble report as the testing is merky at best ... and my brain is mushy at worst.

RJHollins wrote: Is there any way to add 'error checking & reporting' ??

i had a look at the marshal methode and included some error messsaging, maybe it helps you..

oh ... I will look at this right now ...

I have a feeling that my Marshal routine is not put together correctly

I was just now trying to put together a new Save/Load test, and was testing my RUBY module ...

I don't think my data being outputed is coming from the disk save [the container] ... as hard as this is to explain, I think the data I 'thought' was loading ... MIGHT be coming from the Input Arrays of the RUBY module.

No sleep the last few night ... put something doesn't seem right with my doing !!!

NuBeat ... I need some time to look into this latest post you sent. .... IF it was based on my earlier RUBY post ... I may have screwed this all up.

What I'm trying to verify is the data saved is what is loaded and filling the output array. To test this, I enter data ... Save it ... THEN CHANGE the input data again .... then hit LOAD ... the disk data should fill the output array, but my code seems to continue to grab the INPUT data ... NOT the disk data.

RUBY hangs on to all this data internally ... I cannot figure out how to null everything.
I tried:


tricky for me ...

I told you I'm just dangerous with this stuff

Update ...

With what's left of me tonite ...

NuBeat ... thank you very much. I've tested, and tested you Marshal module and all is working !!
I do need to study what you've sent [hope to learn something] !

I have assemble the test module and have it out for my BT. See how that goes

Some interesting developments ... He tested a PRIM based text save and load. I didn't put any auto stuff in there. He had to select a location and name the file. Same for the loading.

We had SUCCESS This is good news. It was an EXE version. still ... this helps to figure what the problem is.

He's now testing the NuBeat 'Marshal' module ... modified slightly for user input.

Will report back with results.

******** Edit *****

Early report back ... using the Marshal Ruby, [by NuBeat], we have written a file to the USER area [APPs] and have it LOAD back in. This is with the EXE form. I think this is the first time this has been done on this test system.

I had my tester send me the saved file ... I was able to load it, BUT I did notice some extraneous characters loading the file He did not mention anything like that on his display. I've asked him to do more testing with this app, and take notice if what's type is displayed exact or not.

When I looked at the saved file in NOTEPAD ... I do see a difference I don't know why ... I would expect things to be exact. That is why I'm letting him test this routine and report.

BTW ... this is using the 'standard' read/write mode of Marshal. I do want the RB version to be tested, but I first want this explored more. Unless he just accidentally bumped a key, don't know.