JP1 Remotes

I recently ran out of space on my remote. I came up with a strategy to free some space that worked pretty well, but has a limitation I am looking for help in overcoming.

My original setup had a macro on each device key, each calling its own LKP, which in turn called one of two macros.

The new setup does away with the macros on the device keys, moves the LKP's from the same phantom key in each device onto the device keys in the AUX device, and incorporates one of the two called macros into the LKP.

After activating the extender, I use a macro (Shift-9) to bind the Other keys to the AUX device, then never bind them to anything else. The problem is that I lose the functionality of the Power key (I have a macro on Scan/Fav, so that one doesn't matter).

I have uploaded both the original and new setups here:
original - http://www.hifi-remote.com/forums/dload ... le_id=4123
new - http://www.hifi-remote.com/forums/dload ... le_id=4124


Is there any way to have the power key linked to a key in the Channel or Menu group, for whatever device is currently bound to that group? Or what about linking the power key to the power key of the device mapped to the last device button pressed? Solutions would have to be less than 56 bytes, or else I'm better off with my original setup.
Also, is there any better (more automatic, or less space) way to bind the AUX device to the Other keys?

An oddity I don't understand is why the LCD displays TV 0000 instead of TV 1000 like it should.
Also, when incorporating the macros into the LKP's I got an error message saying there couldn't be more than 14 total steps, even though I was at 13 or 14. So, I can do 12 steps successfully, when the readme that came with the extender says 13 and the IR message says 14.

If anyone can help me out with any of this, I'd appreciate it.

I am aware (and can think) of no way to reassign the Power button to another button group. I am not sure how much memory you need to reclaim, but you can a save a byte or three here and there throughout your setup. (I am only looking at your original, not your modified IR file)

BTW, outstanding notes and organisation of your IR file. Although what you have done is by no means "easy" to follow, such good annotation really make it easier to unravel.

(1) Do you really need three distinct Pauses defined? Perhaps just define one, and use it 2-3 times when you need an extra-long one. Obviously this may or may NOT save you memory, depending upon how often you use each version of the various-length Pauses;
(2) In your device-selection sequences, I acknowledge the symmetry and "beauty" of having each Short and Long sequence defined as a standalone macro, and then calling those macros in the related LKPs, but you can likely convert at least SOME of those LKPs to having the macro sequence sequence embedded in the LKP, without calling an additional macro. Even when you come up against the 12 (13 or 14?) step LKP limit, you could still define ONE side of the LKP within the LKP and only have the OTHER side called by a standalone macro. I did this with your VCR LKP, i.e. deleted xshift-1 macro, and put that entire sequence in the "Long" side of the VCR-phantom 3 LKP and gained 3 bytes. For your CBL device selection LKP, you can fit BOTH inside the LKP (delete xshift-3 and xshift-4 macros) and gain 6 bytes. So there are 3 or 6 bytes to be gained from each device selection LKP sequence whilst keeping them as universal macros and not resorting to making them LKPs in the AUX device as you have done;
(3) Finally, and obviously somewhat drastically, I as well had an extender setup in which the remaining memory was 3 bytes for keymoves and 0 bytes for upgrades, and wanting to add an additional functionality. After wrestling with my inner demons for a while, I bit the bullet and gave up the Custom Mode Name. It eats a lot of memory (BOTH upgrade and Keymove, since the keymoves are 13 bytes each), and nobody in my house REALLY ever looked at the LCD screen. Your situation may be different, but if you run out of options, that's one to consider. (Or as a minimum, delete those Custom Mode Name keymoves for a few modes, when the room is dark and you're watching a movie perhaps. At 13 bytes a crack, you may only have to live without one or two of those custom labels.)

There may be other possibilities, but the top two may reclaim enough memory for your requirements (again, I don't know whether you need 55 bytes - which is a pretty tall order)

Well, I've spent some more time looking at your IR file (original) and have reclaimed a BUNCH of memory, but also have some questions:

xshift-5 macro - You have this sequence <Dev_AUD;1;DEV_TV;shift-1> twice. If these are input selections, I see no benefit to this, so in addition to moving this to the Long side of SAT:phantom 3, I deleted the first instance of that sequence;

xshift-7 macro - Same basic question <DEV_TV;shift-0;DEV_AUD;4> is executed twice, although the seceond instance of DEV_AUD is covered by the shift-TV macro called in the middle, so I deleted the first instance of this sequence;

Audio:Mute LKP - you have DEV_AUD on each side of this when it can only be called from the AUD device mode already, so I deleted those;

Following the formula I outlined in my post above regarding your device selection sequences, I believe that the attached IR file should behave exactly like your original, yet has 52 bytes of Keymove/Macro spce rather than only 4. I also moved the "ALL OFF" sequence from xshift-Power directly to the long side of CD:xshift-phantom 5 LKP, but don't know if you wanted to retain the ability to call that macro elsewhere.

If I missed some instances of multiple calls to any of these macros, then of course, they won't be called any more, but as they were all on xshifted buttons and seemingly very task-specific, I would imagine that you weren't using them directly nor in other macros.

I also didn't mess with the Pause functions, since I was not inclined to search through every Macro and LKP to determine where they are utilised, but I thingk you have the potential for another 5-10 byte gain there, as I outlined above.

Assuming that every little bit helps, if this upgrade works for your fan, it will save you 4 bytes of upgrade space...

Upgrade Code 0 = 9F CE (CD/1998) Windmere Fan (KM v9.06)
27 00 02 00 79 33 00 62 3A D2 58 C5
End

Upgrade protocol 0 = 01 27 (S3C8+) Windmere Fan Protocol (PB v4.00)
3D 90 21 8B 0E CF 01 08 08 03 B1 01 27 01 3B 03
9D 9D 41 E6 0D 05 E4 05 04 8D 01 46
End

Thanks to you both! You went well beyond the call of duty.

Capn Trips:
I'll look into reusing one pause value.
Moving the long macro into the LKP works well. I noticed you didn't move the DVD macro, did you run into the same ceiling of 12 steps that I did? Is that the correct number, or does something need to be fixed to allow 13 or maybe 14? The short macro is called by the long macro, so needs to stay outside the LKP.
I'll consider getting rid of the CMN's if necessary. At one point I was using them , but sending only the bytes I cared about displaying (<9). Could that do anything bad?
The repeated input selections are meant to decrease response time. It sends them immediately in case the device is already on, then again after the powerup sequence. I used Toadtog to track power status at one point, but I have three 2117's I use from different rooms, so that didn't work well.
Speaking of decreasing response time, the Play and Fast Forward keys aren't as responsive as the original Tivo remote. Would defining them as keymoves and putting them at the top of the list speed up those keys any?

Robman:
That upgrade does work. Thanks. A past suggestion of yours about getting CBL 1190 down to one byte per key instead of two worked well. I don't think there is anything I can do about CBL 1189 since it uses 6 RC6 and 8 RC5 devices.

In using the new setup for a few days, I found the lack of a power key wasn't much of an issue. For now, I plan on using a setup similar to Capn Trips', with the new fan upgrade, but might keep the AUX method in mind for next time I run out of space.

Thanks again.

I just tried it out, and the Audio:Mute LKP doesn't work without the two Dev-Aud's. This reminds me of a problem I ran into. I was going to have the macros on the device keys all call the same LKP on Aux:Phantom3, which would then call either of two DSM's. I tried it, but the LKP wouldn't work.

TV=Dev_TV;Dev_Aux;Phantom3
VCR=Dev_VCR;Dev_Aux;Phantom3

Aux;Phantom3=[Short]Dev_Cancel;Shift-DiscreteOn [Long]Dev_Cancel;Shift-DiscreteOff

where Shift-DiscreteOn and Shift-DiscreteOff are DSM's for each device

I don't understand either one.

mikemcgo wrote: I just tried it out, and the Audio:Mute LKP doesn't work without the two Dev-Aud's.

edited: Hmmm - that's odd. I wonder what's going on there. Oh well, if thats what it takes, then so be it.

mikemcgo wrote: This reminds me of a problem I ran into. I was going to have the macros on the device keys all call the same LKP on Aux:Phantom3, which would then call either of two DSM's. I tried it, but the LKP wouldn't work.

TV=Dev_TV;Dev_Aux;Phantom3
VCR=Dev_VCR;Dev_Aux;Phantom3

Aux;Phantom3=[Short]Dev_Cancel;Shift-DiscreteOn [Long]Dev_Cancel;Shift-DiscreteOff

where Shift-DiscreteOn and Shift-DiscreteOff are DSM's for each device

I don't understand either one.

I suspect you are hoping that the Dev_Cancel only cancels the most recent Dev_Aux, and allows the preceding Dev_TV or Dev_VCR to become active again for the shift-DiscreteON or OFF, right? Dev_Cancel does not work that way. The extender macro sequencing only remembers ONE Dev_xxx command, which is the most recent one, and Dev_Cancel cancles ALL previous Dev_xxx commands in the sequence, so the shift-DiscreteOn/OFF will execute the corresponding function for the previously active device, not anything assigned by a Dev_XXX command.

mikemcgo wrote:Thanks to you both! You went well beyond the call of duty.

Too much spare time on my hands (I gotta get a life!)

mikemcgo wrote: Capn Trips:
I'll look into reusing one pause value.
Moving the long macro into the LKP works well. I noticed you didn't move the DVD macro, did you run into the same ceiling of 12 steps that I did?

Yep. Whenever I got the alert, I kept at least one side's (either short or long) macro. For the DVD, the short side was only 1 command, so i moved that and saved 3 bytes, but the long side was too long. But there were several like that. I similarly couldn't move into the LKP the SHORT sides of VCR, SAT, AUX and CD and the LONG side of AUDIO. Usually the choice of which side I moved into the LKP was arbitrary, as you save exactly 3 bytes either way, but for DVD, the long side alone was too big, and for SAT, I noticed that xshift-6 was also used in another macro, so I left it intact.

mikemcgo wrote: Is that the correct number, or does something need to be fixed to allow 13 or maybe 14?

I remember a discussion in one of the threads many months ago about this inconsistency. The limit is what IR allows, regardless of what the readme's say. If I recall correctly, the 14-key "limit" includes the entire keymove that the LKP generates, (in the SP tab, open one of your LKPs, and click the Hex box to look at the string it generates).

mikemcgo wrote: The short macro is called by the long macro, so needs to stay outside the LKP.

Yeah, I saw that, but the short macro was just 1 command (phantom 1) so I replaced taht in the Long side sequence, as well, allowing deletion of the shortside macro.

mikemcgo wrote: I'll consider getting rid of the CMN's if necessary. At one point I was using them , but sending only the bytes I cared about displaying (<9). Could that do anything bad?

It can not do anything bad, but may give you inconsistent displays for the "unspecified" bytes. They may retain the last displayed character(s), or some randomly-selected one. I remember no consistency, but it does no harm, and may be mildly entertaining - taking bets as to what will be displayed after "DVD" on your LCD screen.

mikemcgo wrote: The repeated input selections are meant to decrease response time. It sends them immediately in case the device is already on, then again after the powerup sequence.

OK, but if you need memory...only you can decide how crucial that .5 seconds is.

mikemcgo wrote:Speaking of decreasing response time, the Play and Fast Forward keys aren't as responsive as the original Tivo remote. Would defining them as keymoves and putting them at the top of the list speed up those keys any?

Not looking at your file right now, but as I recall, you have LKPs on FF and Rew, so your responsiveness will be slowed while the remote waits to determine whether or not you intend to hold the button down long enough to trigger the Long vice Short sequence. Play I don't know - should be OK. Try the keymove and see, I suppose.

EDIT: OOOPS! I just noticed that somehow I deleted your SAT (TIVO) Custom Mode Name Special Function, so re-inserting that and your DEV_AUDs in the Mute LKP yields only 37 bytes remaining free for Keymoves and macros, vice 52.

You still haven't said what you need the extra space for. If it's a device upgrade, then nothing I have done will help. If it's more keymoves (perhaps from a built-in device code) what are they? ... maybe there are some economies to be found there, as well.

Looking over your use of "pauses" I note (perhaps incorrectly) that in your entire setup you only use pauses twice.

(1) In your CD:phantom 3 LKP on the LONG side you use SAT:phantom 5 which is a max pause of $FF;
(2) In your xshift-phantom 2 macro (CBL device power on sequence) you use the AUD:phantom 5 which is a really short pause of $06.

You never use the TV:phantom 5 intermediate length pause ($20) so you could delete that one straightaway and gain 5 bytes.

I agree that it would a LOT of iterations of the $06 to equal the $FF so those two should probably remain separate, although the short one could possibly be worked around without using a formal "pause" but with an unassigned button press that forces the remote to look through the entire list of assigned functions before deciding that there's nothing there.

So you could insert, for example, DEV_TV;xshift-REC;DEV-AUD instead of that pause and save a net of 3 bytes (5 bytes gained by deleting the AUD:phantom 5 pause, but 2 bytes lost by replacing that phantom 5 command with the 3-command sequence above)

Again, the price is a little loss of elegance within your IR file, but if the functionality is the same, does that matter? Many such tricks abound in the world of JP1. No one answer is "right" just whatever works best for you.