The basic problem is that the information on how to map the RMDU info into IRP protocols and parameters is simply not there in a unambiguous, machine readable form. The Cmdtranslator and the DeviceTranslator maps between the "fixed data" and the RMDU user friendly parameters. These are close, but not quite the same as the IRP parameters, It is not that hard to "guess", but this then has to be programmed for each executor. Also the name of the IRP protocol needs considerable work to extract: For example, the executor named "NEC1" can generate IRP protocols NEC1, NEC2, and more ("send 2x then dittos...").
No, The reason why it can import NEC1 (at least in some cases) but not RC-5 is mainly that the hex -> OBC function in the latter case is lsb+complement ("Translator(lsb,comp)"), which I have programmed in (ProtocolsIni.java) , which is not the case for the hex -> obc function for RC-5 "Rc5Translator()". There are 4600 lines of code in com.hifiremote.jp1.translate......
the simple name change of the protocol from what was imported by IRS from RMIR from "RC5" to "RC-5" meant that IRS didn't know what to do with the imported file. I should think that if IRS knew the protocol (executor) by name, it should have access to the proper executor and be able to import the data based on the known executor and thus populate the fields accurately. In this instance, IRS simply pulled in the file and put all the parameters into a single column, and since it could deliver no information on what those parameters were, it offered little in the way of guidance to a working device file.
In the worst case, the the executor is a "recepie" written in a language we do not quite understand, including the meaning of its parameters. If we had an emulator, we could run it and put the result through DecodeIR, would basically solve the problem...Essentially, if I'm understanding, the protocol tells us generally how the data is supposed to look, and the executor is a set of instructions on how each remote is going to be instructed to make our data look that way.
If exporting RMDU updates, the best bet would probably be to implement some sort of generic (e.g. Girr) export facility within RM. This would mean an extension to protocols.ini, for example on how to generate IRP names.
Can you make that into a constructive suggestion?ncoig wrote:It is more curious to me still that when inputting data for a remote, I wanted to save it for later use in the format I saw in IRS. So I exported the parametric remote data as a Grrr, text and a few other formats, but on importing the very same files, in most every instance, the import fails and I have to re-input all the data again. In none of the formats was the import successful such that I could immediately use the file as I had saved it.
All this said, it's a really neat program, I just wish I could exploit more of it!!