Okay, I am stalled on the Sony Stalk front, cuz I cannot figure out exactly how most of you are likely to want this thing to work.

Hijack can now internally intercept/inject "button codes" from the Sony Stalk. The question I have is, how should this capability be made available externally, specifically for [ir_translate] functionality?

I'd like to keep the same or similar syntax as current [ir_translate] functionality, but the Sony Stalk presents unique challenges.

So, you folks who have one, please get familiar with the (much) earlier postings on how the thing sends data (basically, a code for each press, release, twist, untwist, push, and release of all any of the little button/knobby things on the device, all of which can be done in any sequence the user decides to use..).

And let me know what you want, within reason and memory limitations.

Thanks

I don't know how you'd code this up from an internal perspective, but I know how it would be useful to me if I were to use it.

From the end-user point of view, the purpose would be to re-assign functions to the various buttons. For example, if I wanted the SOURCE button on the stalk to, say, switch visuals instead of switching the input source, then I would want to be able to do so. (I don't know if the stalk already has a visual-switch button, I'm just giving an example.)

If, in order to do that, it would be necessary to allow stalk commands to get re-mapped to IR commands, then so be it.

One of the things I'm considering is that my friend Tod is about to get his Chrysler-to-Sony interface for his steering wheel buttons. When he gets it, I don't know whether the built-in functions from the device will do all of the things he wants them to do. If not, it would be nice for Hijack to allow him to make the changes.

Another thing would be to allow Hijack to let other resistor values (other than the Sony values) be mappable. I don't know if that's possible. It would be useful for people who choose to roll their own buttons with handmade resistor networks.

Yes, programmable resistor values will DEFINITELY be there.

And the small buttons are the EASY part. I guess my real question is what to do with combinations of TWISTS and buttons?

Sorry, didn't realize you could get that detailed.

Obviously, the most useful thing would be to give the end-users complete freedom to map any combinations they wanted. The design of the [hijack] entries would, I suppose, be whatever's easiest for you to code and therefore what would be the most likely to get the feature implemented soon.

Although I'm pretty happy with the stalk setup now (althoughy my button presses are still kind of temperature dependant) a quick voladj toggle (Low.Med.High.Off.Low etc) would be nice.

If full stalk button remapping is implemented a little piece of front end config.ini editing software would be nice, as frankly all the options available would make editing the config.ini confusing / error prone for schmoes like me.

What I imagine is this: three columns with rows of combo boxes. The columns would be 'Action1', 'Action2', 'Empeg Command'. The output would be little text file you could cut & paste into your config.ini. So if you wanted a button press/twist combo the button press would be chosen in Action1 and the twist in Action2 and the desired action chosen in Command.

Premature, I know. But if the capability to completely reprogram the stalk existed a tool like this would make it very easy to goof around with the settings & easier for Lazy, Overworked Nonprogramming Comfortable Userland Joes like me to fiddle with it.

-Zeke

there are 2 areas here.

1. Allowing users with no Stalk control to do maps so that selected IR/front panel buttons are mapped to regular Stalk buttons for things like Visual- - these can hopefully passed to the player to action as fake stalk codes.

2. For those of use with a stalk, allow us to define what the buttons are mapped to within the player and possibly define new button sequences and generall make the Stalk do all sorts of things.

I am mainly interested in the latter as I have a stalk and therefore being able to go visual- is something I can do now. However this would be useful when at home when selecting the visual to display as often it quicker to go back than "go around the horn again".

So, for my input to point 2, the Sony Stalk hardware lets you do two main things right now:

1. Press a button, twist the volume knob or the seek ring forward/back - but you can only do one of these 3 things at a time. There is no concept of chording (pushing mutiple buttons at the same time) on the stalk due to its design, except for the volume knob being pushed in along with another button.

2. Press and Hold the volume knob in, (this activates the shift function), then you can twist the now locked together volume knob/seek ring forward or back or press another button.

Note: other than seek+/- which does next/previous visual the other shifted buttons are ignored by the player right now so I guess the player knows they are shifted button codes, its just that theres no function assigned to them so it ignores them.


What I would like Hijack to be able to do is to trap all the unshifted buttons as 'stalk codes' (fake IR codes or symbolic stalk names), these same buttons can all be pressed and Held like regular IR buttons can (so the .L button maps are allowed for).

The only thing here is that what button is volume up/down depends on the stalk setting in Emplode (which is in the config.ini), I don't know if it will be easy to allow users maps to talk about StalkVolUp say as a symbolic button and have hijack work out whether that is really the vol+ button or actually the stalk vol- button due to it being mounted on the left rather than the right. (this would let you and me have the same 'logical' stalk map even if I have mine on the right hand side of my steering wheel and you have yours on the left.

Comments on this Mark?

As well, I would like the Shift function handled as both a 'shift modifier' so that all button codes you press while Shift is held in now report/can be mapped as shifted versions of themselves. (.S modifier)

And I would also like the quick push in and then release (with no other button involved) of the volume knob handled as a special 'modifier' button so that for example:
a push and release of the shift button followed by the push of the Off button is seen as a distinct/different from just a push of the off button, and is also different from shift in (and hold) and then push of the Off button.

This shift button as temporary modifier option lets us get at 8+ more assignable functions from the same stalk control, its probably optional but would be a nice to have - could be useful to assign some useful codes to it as it can be done with one finger without looking (push the knob in then release to activate modifier, then press other button), these buttons would be nearly as useful as regular unshifted buttons

I would expect a time out of say 1-2 seconds for the push and release modifier option (i.e. if you don't touch another button within 1-2 seconds of the pusha dn release the shift button then the modifier state is cancelled). Also once any key is pressed the modifier state is cancelled so its a single button press option.

The push the knob in (shift & hold) then pressing another button is a normal shift function - I believe all buttons can be shifted (except the seek+/seek- vol+/- which reports as shift vol+/vol- becuase they lock together when shifted).

Hope that explains what I want it to do.

With the unshifted/shifted codes we can get something like 16 buttons available, with the shift in and release as an additional modifier we get another 9, that gives us more buttons on the stalk than the regular Ir remote supports, and since the Shift and unshiefted could have long presses as well then thats lots and lots of options from one tiny stalk [gives rise to lots of 'macro' buttons we could have to do whizzy things that currently require many button presses].

Of course thats my full wish list with Stalks, how much Mark can actually implement is another story.