Is this me, or is it an RDF problem?

daveoram · Post by **daveoram** » Wed Mar 03, 2004 2:45 pm

I am not sure if it is me doing something wrong or if there is a problem with the RDF file.

I am using RM 0.94 (just started using instead of KM) and latest RDF's to try and program an upgrade for my 8550.

It works OK, but seems to map the transport keys to the wrong keys (ie the PLAY function is mapped to the REWIND key, PAUSE maps to STOP, ENTER maps to PLAY amongst others.

I have posted my RM file here if it helps explain things

I am sure it is me doing something wrong, unless it is the button assignments which are different in the 5550 and 8550 and the RDF is supposed to be universal.

Give me some pointers and I will try and sort the RDF out - I know this is not a popular remote, but I regard it as a challenge to try and sort this.

Dave

Nils_Ekberg · Post by **Nils_Ekberg** » Wed Mar 03, 2004 3:45 pm

I did a quick test using your RM upgrade and the first time I did it I got the same condition. I then looked closer when I did the add and realized it forced me to pick a device type and I had picked the wrong one. Since it is a DVD upgrade you need to make sure you pick DVD and it will map correctly.

IR assignes the keycodes when you add an upgrade based on the order of the button map for the chosen device and if you create the upgrade with one button map then pick a different device type in IR you are actually picking a different button map so the buttons get assigned in a different order.

Mark Pierson · Post by **Mark Pierson** » Wed Mar 03, 2004 3:57 pm

I find it odd that RM is treating the transport keys as keymoves, but IR shows them mapped in the upgrade (with different hex codes).

I just re-created the upgrade in KM and come up with vastly different code which appears to be mapped properly in IR:

Upgrade Code 0 = 93 E9 (DVD/1001)
4B 06 3E FA FF B7 60 5C 88 18 08 30 7C 80 78 9C
6C 98 64 2C 28 24 20 90 08
KeyMoves
B2 F3 93 E9 90
End

Nils_Ekberg · Post by **Nils_Ekberg** » Wed Mar 03, 2004 4:30 pm

I thought it was a little odd also that I had to pick the device type (DVD) to match the RM output with Dave's upgrade. I went back and took one of my other upgrades in RM for a TV and changed it to the 8550 and loaded it into IR with no trouble. I then took one of my DVD upgrades for another remote, converted to the 8550 and loaded it with the same problem of it not assigning the device type.

It seems atleast for this remote and DVD RM may not be packaging the DVD updgrades properly.

gfb107 · Post by **gfb107** » Wed Mar 03, 2004 6:03 pm

The transport keys are not in the keymap for DVD because of an an error in the RDFs, as documented in this post.

Using that modified RDF, the generated upgrade is:

Code: Select all

Upgrade code 0 = f3 e9 (DVD/1001)
 4b 06 3e fe fd b4
 60 5c 88 18 08 30 7c 80 78 9c 6c a8 98 64 2c 28
 24 20 90
END

The upgrades are still different, but not vastly so. Looks like they must be using different ButtonMaps. I assume IR is using the same one as KM.

My suggested changes to the RDF must be incorrect, at least in the case of the DVD DeviceTypeAlias. I wish I knew where to look in KM to find the ButtonMap information so I could do a sanity check. Well, I wish I did, but I certainly don't have time to learn how KM works, I don't really have the time to do what I do with RM.

gfb107 · Post by **gfb107** » Wed Mar 03, 2004 6:56 pm

OK, now I'm even more confused. I recreated the upgrade in KM 8.16 (get it here. This was rather hard to do because of the strange button names used in the 8550/5550 RDF.

Here's what KM generated:

Code: Select all

Upgrade Code 0 = 93 E9  (DVD/1001)
 4B 06 3E FE FD BF 60 5C 88 18 08 30 7C 80 78 9C
 6C A8 98 64 2C 28 24 20 90                     
End

There are 2 differences I see from what KM generated.

First is in the upgrade header, where KM generated 93 E9 (DVD/1001) where RM generated f3 e9 (DVD/1001). I can't remember right now what that difference means.

The second difference I see is in mask. KM generates 3E FE FD BF while RM generates 3e fe fd b4 I think the difference is that KM fills left-over bits in the mask with 1's, but RM uses 0's. This isn't supposed to make any difference.

daveoram, can you test the above KM generated upgrade and see if it works any better for you?

Post by **The Robman** » Wed Mar 03, 2004 7:22 pm

gfb107 wrote:I wish I knew where to look in KM to find the ButtonMap information so I could do a sanity check.

If you mean which buttons are included for each device mode for the remote in question, all you have to do is look at the "Layout" sheet, that's where the info is displayed.

gfb107 · Post by **gfb107** » Wed Mar 03, 2004 7:27 pm

OK, That first digit is supposed to be the device type index, which among other things, identifies the ButtonMap being used.

So KM thinks the device type index is 9, while RM thinks it is 15.

I can see in the RDF why RM chose 15. I also see now why the DeviceTypeAliases section was written the way it was, although it caused the wrong ButtonMap to be used, it did at lease result in RM and IR using the same device type index.

I'm going to need help from the real experts to figure out how to fix this RDF. The problem lies in either the DeviceTypes or DeviceTypeAliases sections (or both).

It looks to me like we've got another RDF problem.

gfb107 · Post by **gfb107** » Wed Mar 03, 2004 7:31 pm

The Robman wrote:
gfb107 wrote:I wish I knew where to look in KM to find the ButtonMap information so I could do a sanity check.
If you mean which buttons are included for each device mode for the remote in question, all you have to do is look at the "Layout" sheet, that's where the info is displayed.

Thanks Rob, but I was looking for more detail than that. Something that would be roughly equivalent to the [DeviceTypes] and [ButtonMaps] sections in the RDF

Fortunately, I only wanted that information because I was chasing a wild goose.

Mark Pierson · Post by **Mark Pierson** » Wed Mar 03, 2004 7:53 pm

gfb107 wrote:Something that would be roughly equivalent to the [DeviceTypes] and [ButtonMaps] sections in the RDF

[DeviceTypes] = Dev Types
[ButtonMaps]= This Dev (with the Topline as the selected remote)

The problem lies in either the DeviceTypes or DeviceTypeAliases sections (or both).

I think the problem lies in here somewhere:

Code: Select all

[DeviceButtons]
SAT	 = $026 $01E $02E
TV	 = $027 $01F $02F
VCR	 = $028 $020 $030
CD	 = $029 $021 $031
@AMP	 = $02A $022 $032
@TUNER = $02B $023 $033
AUX1	 = $02C $024 $034
@AUX2	 = $02D $025 $035

[DeviceTypes]
SAT        = 0
TV         = 0
VCR        = 2
CD         = 3
@AMP       = 4
@TUNER     = 4
MISC_AUDIO = 4
TAPE       = 6, $070D
VID_ACC    = 0, $0008
CABLE      = 1, $000C
PHONO      = 3, $030B
AMP        = 4, $0604
TUNER      = 4, $0605
HOME_CT    = 6, $060A
LASER      = 5, $0707
DVD        = 3, $0709
DAT        = 6, $070E

From "DVD=3, $0709", KM uses DeviceIndex=09, ButtonMap=3.

The Topline has been in KM for a long time, and I don't recall there ever being a problem such as this before.

gfb107 · Post by **gfb107** » Wed Mar 03, 2004 8:13 pm

Yep, you've spotted it. I didn't realize that the low-order byte of the type determined the device type index. I thought it was strictly positional. I thought I read that somewhere when I first started working on RM ( back when we called it KMJ, and sometime called KM KMX (X for eXcel ) to distinguish the two).
I'll fix the code in RM that parses this part of the RDF, of course.

gfb107 · Post by **gfb107** » Wed Mar 03, 2004 9:17 pm

OK, RM v0.95 is now available, and it fixes this problem.

daveoram · Post by **daveoram** » Thu Mar 04, 2004 3:41 am

Sorry I've not got back to you all earlier, but I wil test both Gregs code and the new version of RM as soon as I get back from taking the kids to school - its a lot easier to concentrate/work without 6 little terrors trying to 'Help'.

Dave

daveoram · Post by **daveoram** » Thu Mar 04, 2004 5:45 am

No guys, still no joy.
I will post a table showing exactly what is programmed onto each button, and what is actually sent to each button once the upgrade file is sent to the remote.
It may take a while to do though

Dave

Mark Pierson · Post by **Mark Pierson** » Thu Mar 04, 2004 6:14 am

My guess is the DeviceTypeAliases are still incorrect in the RDF. I unfortunately don't have time to look into it now, so maybe one of the others can help.