Using LKP instead of shift.

gfb107 · Post by **gfb107** » Wed Jan 26, 2005 7:04 pm

I've recently started using LKPs, and have found that I prefer them over shifting. However, they take a bunch more work to define.

What I would like to see is an enhanced extender that had, as an option, the ability to simply turn a long key press into a shift. That way a short press would send the "normal" command, and a long press would simply send the "shifted" command. If nothing is defined on the "shifted" key, it should just continue sending the "normal" command (or the last command of a macro or DSM).

This would work really well when you want to just issue a single key for "short" and different key for "long". If a sequence of commands is desired, the user would create a macro or DSM on the "normal" or "shifted" key.

To get what I want, in the new extender I would

Enable "Use long press to shift" in the "Other Settings"
For each LKP
1. Assign the desired long function to the "shifted" button (no keymove required, could be defined as part of the device upgrade/setup code)

To accomplish the same thing today, I have to:

For each LKP
1. Assign the "normal" function I want to the shifted button.
2. Assign the "long" function I want to the xshifted button
3. Create an LKP assigned the the "normal" button that invokes the shifted button for the short press, and the xshifted button for the long press.

The draw-back I see is that LKP is not as effective at hiding commands that you don't want accidentally pressed. Of course, xShift would still be available for that purpose.

Thoughts?

Nils_Ekberg · Post by **Nils_Ekberg** » Wed Jan 26, 2005 7:53 pm

Greg, I think that is definitely doable. What should it be called? How is "Auto Shift"

The simplest approach would be to make a modified version of the LKP protocol that when placed on a button executes the normal function for a short and whatever is assigned to shift for the long.

If I can limit it to just one code on each it may be pretty quick to put together. It would still need at least one parm for duration since there is no one good default for all remotes.

Which remote should I target first?

Unless someone else volunteers tonight I will look at it tomorrow.

gfb107 · Post by **gfb107** » Wed Jan 26, 2005 8:13 pm

Nils,

Thanks for your interest.

I was actually thinking that it should NOT be a protocol, but instead should be part of the extender itself, much like the "virtual" TV/1103 device in the URC-6131 extender for DSM. The only thing I want to have to do to enable this "auto-shift" behavior is set it in the Other Settings area on the General Tab in IR. If I have to create a keymove for each key I want it on, we haven't really changed the effort required to use LKP. Once "auto-shift" is enabled, I want it to apply to all keys in all device modes, regardless of setup code.

I suppose if you wanted to get fancy you could create a flag for each device mode, or even each key set, like some remotes have the ability to set different VPT devices for each device mode.

In terms of which remote to target first, I personally use URC-6131s and URC-7800s, so those would be the ones I could test on. I should be getting one of Rob's "certified pre-owned" URC-8910s in the near future so I could test on that later too.

BTW, I've edited the first post to make that clearer (hopefully).

Nils_Ekberg · Post by **Nils_Ekberg** » Wed Jan 26, 2005 9:07 pm

Let me think this out a little. The only real problem that comes to mind if it is global it would also eliminate "key repeats" like channel up and down and volume, etc.

gfb107 · Post by **gfb107** » Wed Jan 26, 2005 9:51 pm

Repeats would still have to work. If there isn't anything mapped (setup code, keymove (including DSM), or macro) to the shifted button, it should repeat the unshifted button (which could also be from the setup code, keymove (including DSM), or macro). If there is something mapped to the shifted button, that should repeat. If a macro or DSM is being repeated, the last key should repeat.

e34m5 · Post by **e34m5** » Thu Jan 27, 2005 7:25 am

Sounds to me like this should be a part of the macro page in IR. As Nils said what is required is simply a check box called "auto LKP".

By definition the short portion would be the default behavior of the button and then the long would be whatever the macro definition is.

Nils would not have to change anything in the RDF's. This is all internal to IR.

gfb107 · Post by **gfb107** » Thu Jan 27, 2005 7:33 am

But what if there isn't a macro on the shifted key? What if it is just a keymove? Or even defined in a setup code/device ugrade (some remotes have some shifted keys in some of their keymaps)?

e34m5 · Post by **e34m5** » Thu Jan 27, 2005 7:44 am

Not following. I'm suggesting that there would be 4 variants for macros:

1 - Normal
2 - Shifted
3 - X-Shift
4 - Auto LKP

So that each button could have up to 4 macros defined. That may be more than we need. Perhaps the variants should be exclusive....not sure how I feel about that.

As is the case now category 1 totally redefines the function of the button press. However for the new cat 4 it would behave just like 2 and 3 do now, which is simply that a normal button press performs the default behavior.

However, maybe this needs to be an option in the SP page. Maybe there is a selection when creating LKP's thats says "default" this could be used either on the short or long side of the LKP.

gfb107 · Post by **gfb107** » Thu Jan 27, 2005 8:33 am

My point is that I want "auto LKP" to be entirely automatic and global. Not on a macro-by-macro or keymove-by-keymove basis. Every button that has anything defined on the shifted-key (most of which would be generated as keymoves as part of importing a device upgrade) would automatically have LKP behavior.

Let's say in my device upgrade (as seen in KM or RM) there are 10 functions mapped onto shifted buttons, which results in 10 keymoves onto the shifted button in IR. I don't want to have to do anything to those keymoves in IR, I just want them all to be invoked by LKP. I don't want to have to create new macros, or modify the keymoves in any way.

e34m5 · Post by **e34m5** » Thu Jan 27, 2005 8:39 am

So you don't want Shifted functionality any more? Perhaps KM needs to be modified to flag certain keymoves as "auto LKP"

gfb107 · Post by **gfb107** » Thu Jan 27, 2005 8:51 am

No, I want LKP to be an alternate way to shift. Shift doesn't need to stop working. I don't want to have to do anything in KM/RM/IR to get this behavior, other than setting a global flag in the "Other Settings" area of the General Tab in IR to enable this function. I would call it "Use LKP as shift", and it could be either On or Off, much like VPT.

There should be no changes to IR/RM/KM. The change could be made entirely in a new extender, which would recognize LKP and treat it as shift. A new RDF to go along with this extender would make the "Use LKP as shift" setting appear in the "Other Setttings".

e34m5 · Post by **e34m5** » Thu Jan 27, 2005 8:54 am

I wouldn't do it globally because I may still want some to be Shifted and some to be LKP'd.

I think just a check box in the SP dialog would suffice.

gfb107 · Post by **gfb107** » Thu Jan 27, 2005 9:43 am

I understand that not everyone would want it to be global.
Those of us who want it globally would install the extender that implements this.

Those of use who don't want it globally are probably OK with the current LKP implementation. Use the Special Functions in IR to setup exactly which keys have LKP and what they do for short and long.

For the way I want to use my remotes, that is cumbersome.
I have to do all of the following for every key where I want LKP as shift:

Create a keymove to put my desired "short/normal" function on "shifted" button
Create a keymove to put my desired "long" function on the "xShifted" button.
Create an LKP on the "normal" button that invokes the "shifted" button for a short press and the "xshifted" button for a long press.

Not only is this a lot of work (say I have a total of 20 buttons where I would use this out of the 8 device I currently control, which is about 3 for each device mode), I also end up using a lot more keymove space because I have 2 more keymoves for each LKP (1 to put the normal function on a different key, and 1 for the LKP itself) Assuming single-byte commands, that's 5 bytes for the keymove and 7 bytes for the LKP, for a total of 12 bytes. That's 240 bytes for all 20 buttons. That's a lot when all I want is an alternate method to shift.
With my proposal, all I have to do is

Assign my desired long function to the "shifted" button. I can do this in IR, or as part of my device upgrade in RM/KM

.

e34m5 · Post by **e34m5** » Thu Jan 27, 2005 9:54 am

Agreed.

Now it's up to the IR developer to figure this out.

gfb107 · Post by **gfb107** » Thu Jan 27, 2005 10:08 am

The way I proposed it, no changes are required to IR/RM/KM. We need a new extender that makes LKP detection part of normal keystroke processing, (as opposed to having it in the LKP protocol) with an associated RDF that has the changes necessary to cause IR to show the "Enable LKP as shift" option in the "Other Settings". The RDF might even change the "Shift" label to "Shift/Long", which would then be reflected in IR and RM.