GUI

PhracturedBlue wrote: I created a 'Fixed' curve that can be used for creating constants. Use 'offset' to move it up and down

Thanks, but for example on the Expo&DR mixer there is no offset, but there is scale...

FDR wrote:

PhracturedBlue wrote: I created a 'Fixed' curve that can be used for creating constants. Use 'offset' to move it up and down

Thanks, but for example on the Expo&DR mixer there is no offset, but there is scale...

I see. I couldmake it zero, min or max, but instead maybe we just need to show offset on the expo page.

PhracturedBlue wrote: I see. I couldmake it zero, min or max, but instead maybe we just need to show offset on the expo page.

Well, there is not much room!
Furthermore, I like that page very much...

PhracturedBlue wrote: I see. I couldmake it zero, min or max, but instead maybe we just need to show offset on the expo page.

I think two functions min and max would be great. They could be made zero with the scale if needed, but the point is zero needed much rarely...

FDR wrote:

PhracturedBlue wrote: I see. I couldmake it zero, min or max, but instead maybe we just need to show offset on the expo page.

I think two functions min and max would be great. They could be made zero with the scale if needed, but the point is zero needed much rarely...

Well, i took it one further. There is still only 'Fixed' but it is equivalent to your definition of 'max'. Scale will move it from 100 to -100, so you can access it from the expo page.

Cool!
A bit unusual, but cool...

A suggestion: it would be useful, if it was visible on the mixer list which channel is reversed. Not the inverted source, because it is there, but the reversed channel output...

...and I think I was wrong, when I forced to scroll the mixer page by one!

After all, with this scrolling speed, it would be better to use simple paging...

FDR wrote: ...and I think I was wrong, when I forced to scroll the mixer page by one!

After all, with this scrolling speed, it would be better to use simple paging...

you don't like that long-press on the scrollbar buttons moves it down 1 page?

FDR wrote: A suggestion: it would be useful, if it was visible on the mixer list which channel is reversed. Not the inverted source, because it is there, but the reversed channel output...

See what you think of it now.
Note that the graph will now also be affected by the channel reverse. I'm not sure this is the right thing to do though. Generally you use channel-reverse when the servo direction is wrong. I'm not sure it makes sense to show it on the xy-graph. But I'm willing to go either way on it.

PhracturedBlue wrote:

FDR wrote: ...and I think I was wrong, when I forced to scroll the mixer page by one!

After all, with this scrolling speed, it would be better to use simple paging...

you don't like that long-press on the scrollbar buttons moves it down 1 page?

I mean scrolling with the pencil. When you use the touch screen you would not like to pushing buttons...

PhracturedBlue wrote:

FDR wrote: A suggestion: it would be useful, if it was visible on the mixer list which channel is reversed. Not the inverted source, because it is there, but the reversed channel output...

See what you think of it now.
Note that the graph will now also be affected by the channel reverse. I'm not sure this is the right thing to do though. Generally you use channel-reverse when the servo direction is wrong. I'm not sure it makes sense to show it on the xy-graph. But I'm willing to go either way on it.

I don't think it should be reflected on the graphs, because it is (as you said) for correcting a wrong servo direction only!

FDR wrote: I mean scrolling with the pencil. When you use the touch screen you would not like to pushing buttons...

A long-press with the pencil should also page down...

PhracturedBlue wrote: A long-press with the pencil should also page down...

Yep, you are right...

I tried to setup timers to start with throttle.
If I choosed THR (i.e. throttle stick) it starts, but it doesn't stop when I hit the throttle hold, so I have changed the switch to Ch3. To my surprise it doesn't stop on throttle hold either, in turn it does on the channel output page...

(EDIT: I have implemented the throttle hold with the safety feature on the channel page. It is much more easier, than defining a plus complex mixer page, not to mention the bug, that ruins the complete channel, when I trying to add an other complex page to it...)

I have already mentioned that there are cases, when after some setting changes it doesn't save them, or even messes up thing which were not touched, but I cannot reliably reproduce it!

Last time it was my CB100 model data: I have changed back the aileron and rudder reverse, and after it all the channel mixers were gone. (They all were simple mixers with expo function...)

One example which I can reproduce is when in my GCP model data (attached below) I try to add a forth complex page to the throttle channel (ch3), and implement a throttle hold with it: switch=RUD DR, mux=replace, curve=3point, offset=-100.
After I click OK, the THR disappears from the mixer list page, and opening it again shows a complex mixer with only one page with some defaults.

This is the most annoying bug in the current fw, it is what I used to call, that "it is not reliable".

Please do something!

FDR wrote: I tried to setup timers to start with throttle.
If I choosed THR (i.e. throttle stick) it starts, but it doesn't stop when I hit the throttle hold, so I have changed the switch to Ch3. To my surprise it doesn't stop on throttle hold either, in turn it does on the channel output page...

The rule is currently that the timer will run if the channel is > -80%
There was a bug in how using output-channels worked though. I've fixed that now so you should be able to use channel-3 as a timer input

One example which I can reproduce is when in my GCP model data (attached below) I try to add a forth complex page to the throttle channel (ch3), and implement a throttle hold with it: switch=RUD DR, mux=replace, curve=3point, offset=-100.
After I click OK, the THR disappears from the mixer list page, and opening it again shows a complex mixer with only one page with some defaults.

The problem here was that I had only allocated 16 mixers. You were using all of them, and then treid to add one more. The code throws a message in the emulator but not on the Tx (I don't have an error-message system yet). Additionally, it erased the old mixers before erroring out.

a) there are now 48 mixers which should keep you from hitting the limit
b) The tx will error out before clearing the mixers, so while you won't get an error telling you that it failed, at least it won't erase your previous work.

PhracturedBlue wrote: The problem here was that I had only allocated 16 mixers. You were using all of them, and then treid to add one more. The code throws a message in the emulator but not on the Tx (I don't have an error-message system yet). Additionally, it erased the old mixers before erroring out.

a) there are now 48 mixers which should keep you from hitting the limit
b) The tx will error out before clearing the mixers, so while you won't get an error telling you that it failed, at least it won't erase your previous work.

Cool, thanks!

But I guess it was not the case with the other glicth of the CB100, since it had only 4 simple mixer...

I'm thinking that the channel-reverse should be applied at the protocol rather than at the mixer.
Currently, if you reverse a channel and then display it on the main page (or use it for a timer) it uses the reverse value in the calculation.
However, the purpose of the Reverse is to handle reversed servos, so this seems to be wrong.
But what about the other properties?
Subtrim is basically the same. the purpose of subtrim is to zero the servo, so if you need 5% subtrim, you don't want the value to show 5%, since the purpose was to set the servo to '0%'
How about limits? if we limit the throw to 80% should the max value be 80% Again, I'd think the answer is no. The limit should probably be a final-scaling value (rather than a cuto-off). so a value shown of 100% on the screen would be scaled to 80% (or 125% or whatever) of servo throw.

Do I have this right?


EDIT: However, the safety SHOULD be used when calculating what to show on the screen and for the timers

Absolutely agree with all.