RMIR v2.03 A28 B19 kills prefix device and command fields

jeajea · Post by **jeajea** » Wed Apr 08, 2015 3:57 pm

When you us RMIR v2.03 Alpha 28 build 19 to load a device upgrade with
protocol : Pioneer MIX the “Prefix device” and “Prefix command” fields are
set to none.

This make many buttons in the upgrade fail to send the correct signal.

RMIR v202a doesn’t have this problem

This is the file before loading into 2.03

https://www.hifi-remote.com/forums/dload ... e_id=13309

Post by **ElizabethD** » Wed Apr 08, 2015 7:38 pm

Jim,
I suspect this might be related to the oldsports thread
https://www.hifi-remote.com/forums/viewtopic.php?t=16006
see last few posts. Once fixed up, build or earlier (March 10-17) showed parameters correctly and I see them correctly now.

I just opened the file you posted, and on the function sheet most parameters are null. Is that what you expected? Do you have a file that's older than that one, or perhaps a KM file?

jeajea · Post by **jeajea** » Thu Apr 09, 2015 4:16 am

The file I posted has prefix device and prefix command on most of the
functions as can be seen if you load it in v202a. if you load it in 203 they
are all set to none.

Post by **ElizabethD** » Thu Apr 09, 2015 8:54 am

Try this. Open your .rmdu file. Copy what's on the output sheet.
In RMIR, on Devices tab click New, and Import Raw. Paste the ouput in the upper box and OK. It should work because output data has fixed bytes included, not zeroed out.
Compare parameters on the setup sheet to what you liked in 2.02a. Fix if needed.

jeajea · Post by **jeajea** » Thu Apr 09, 2015 9:16 am

I opened the file in 202a and copied the output and shut down 202a. I
started 203 and did the import raw and the prefix device and prefix
command are still “none”.

203 has a problem with Pioneer mix.

I am running JP1.3 extender 3.03 which needs RMIR 203.

My workaround is to use a second remote to control this DVD player.

Post by **mathdon** » Thu Apr 09, 2015 11:47 am

This one has taken a great deal of tracking down. The bug appears to have been present in every development version since v2.02a and is specific to protocol variant pid 007E:3. I believe it to be a bug in code by 3FG, so will ask him to check this, but I have corrected what appears to have been an error in his code and have posted the jar file for RMIR v2.03 Alpha 28 build 22. Please replace the jar file in build 19 with this one and test if now all seems well.

For 3FG: Please look at PioneerMixTranslator.java. I have made the following amendment:

Code: Select all

case 2: // Device
            if (execVariant == 3) {
                if  ( val == 0 )       //was (val < 4), changed 9 April 2015 by GD as apparently an error
                  flag &= 0xFE;
                else 
                  flag |= 1;

to code at line 68. I have put the amended code in the SVN. If I've got this wrong, please make any necessary corrections.

jeajea · Post by **jeajea** » Thu Apr 09, 2015 12:57 pm

Thanks
It looks right when I open it for edit and the buttons that weren’t working
before now work.

Good job (and quick to)

3FG · Post by **3FG** » Thu Apr 09, 2015 9:21 pm

Graham,
Yes, I agree with your change. Jim, I'm sorry about causing trouble!

Post by **ElizabethD** » Fri Apr 10, 2015 10:55 am

Graham, 3FG,
Could you also take a look into the oldsports thread I mentioned
https://www.hifi-remote.com/forums/viewtopic.php?t=16006
because I think build 22, for variant 4, still might need some change, other than what I did there.
It's not mine, and I don't need it, but I was curious if anything changed, and I don't think so.

3FG · Post by **3FG** » Fri Apr 10, 2015 9:11 pm

Liz,
I don't see anything wrong with Variant 4, but I could easily miss it. Do you have an example of the problem?

Post by **ElizabethD** » Sat Apr 11, 2015 2:46 pm

Well, the example is in the thread I mentioned. My best summary of that thread is like this:
When oldsports took KM file for RCU810, put it into RM, and changed remote to Atlas, parameter things got wiped. I then made the RCU to Atlas conversion in KM, and when that was loaded into RM, it worked fine with a bit of rearranging of parameters on the RM setup sheet.

Additional information
I saw some variant difference for 2 and 4, a subject well over my head:
When I load KM RCU file into RM, and save as RMDU it lists variant 2

Protocol=00 7E
Protocol.name=Pioneer MIX
Protocol.variantName=2
ProtocolParms=166 175 161 160
FixedData=9A 7A 0A FA

When I load KM RCU file into RM, change the remote to Atlas and save as RMDU, it lists variant 4

Protocol=00 7E
Protocol.name=Pioneer MIX
Protocol.variantName=4
ProtocolParms=null null null null null null
FixedData=FF FF FF FF FF FF

and the conversion drops the expected data.

So it looked to me like KM is still a necessary tool

3FG · Post by **3FG** » Sun Apr 12, 2015 12:57 am

Liz,
I added a post to the other thread which is intended to explain what is going on there, and to describe how to make the conversion using only RM.

I can't guarantee that PioneerMixTranslator.java is entirely correct, but I think the cause of oldsports' issues with conversion is more fundamental.

FWIW, as I try to explain in the other thread, I don't see KM as necessary or even desirable for conversion.

Post by **mathdon** » Tue Apr 14, 2015 11:39 am

3FG wrote:I don't see anything wrong with Variant 4, but I could easily miss it.
...
I added a post to the other thread which is intended to explain what is going on there, and to describe how to make the conversion using only RM.

Although there is nothing wrong with variant 4, I have made a change to it that improves the handling by RM of the conversion from variant 2 to variant 4. See this post in the other thread. As before, I would like 3FG's approval (or otherwise) of this change.