Re: Junk string in ly -> midi -> ly conversion

[Mojca Miklavec]

On 20 July 2016 at 18:08, Federico Bruni wrote:
> Il giorno mer 20 lug 2016 alle 17:51, Mojca Miklavec ha scritto:
>>
>> I tried to compile lilypond (2.19.44) from source. And the midi file
>> generate with that binary seems OK.
>>
>> (I'm even more happy that the binary didn't actually crash because
>> it's compiled against a different stdlib than its dependencies.)
>>
>> Is it possible that the problem with the official binary is only there
>> because I'm using a 32-bit binary on a 64-bit system?
>
> Maybe this issue explains it?
> https://sourceforge.net/p/testlilyissues/issues/2208/

This "explains" why I have to jump through hoops to be able to run
midi2ly as a 32-bit python 2 script.

But it doesn't explain why the midi file generated by lilypond is buggy.

------------

The problem in the above link would be addressed by providing 64-bit binaries.
I'm staring at
    http://lilypond.org/gub/
    https://github.com/gperciva/gub
but cannot quite understand why 64-bit darwin isn't supported except
if the following sentence from README is still true (but it would be
absurd if it was indeed still the case):

      • Hourly build of LilyPond subpackage for all platforms except
native (native
        is darwin-pcc?)

(If the machine where the binaries are built is indeed a PPC machine,
I wonder if endianess is screwing things up :) :) :)

Mojca