[L2Ork-dev] updating cyclone

katja katjavetter at gmail.com
Thu Mar 30 23:41:14 UTC 2017


On Thu, Mar 30, 2017 at 8:46 PM, Jonathan Wilkes <jancsika at yahoo.com> wrote:
>> It must be a makefile issue after all, in the library root makefile.
>
>> Pd-lib-builder doesn't define PD_INCLUDE. It does define PD_PATH for
> PD-X compatibility and should find binary and source files from there
> if the tree layout is similar. If not, you can specify PDBINDIR and
> PDINCLUDEDIR in the central library makefile.
>
> You can see from
> http://disis.music.vt.edu/pipermail/l2ork-dev/2017-March/001235.html
>
> that the first line is feeding in a sane location for PD_PATH. The tree
> structure is exactly the same as pd-extended.  That looks like sane
> input to me.
>
> But this does not look like sane output to the linker flag:
> --enable-auto-import "/pd.dll"
>
> There's no reason for "pd.dll" to exist in the toplevel directory of the
> build environment.

The path for pd.dll is missing because it wasn't found. If you have
'PD_PATH=/home/USER/purr-data/pd', pd-lib-builder will try to find
header files in '/home/USER/purr-data/pd/src/' and pd.dll in
'/home/USER/purr-data/pd/bin/'. If your tree layout doesn't have
source and binary directories under the same parent, don't use the
PD_PATH variable but specify PDINCLUDEDIR and PDBINDIR.

Katja


>
> -Jonathan
>
>
>
>
>
>> Katja
>
>
>
> On Thu, Mar 30, 2017 at 6:46 PM, Alexandre Torres Porres
> <porres at gmail.com> wrote:
>>
>> yeah, I thought it might be something in pd-lib-builder
>>
>> but I just tested it again here on windows (finally had my hands on the
>> machine) and it was all fine. I'm using MINGW32, the pd-lib-builder, and I
>> have the latest vanilla in the proper folder, so I guess it finds what it
>> needs to without any hassle
>>
>> you probably should be able to edit things and make it find what you need
>> anyway
>>
>> cheers
>>
>> 2017-03-30 12:55 GMT-03:00 Jonathan Wilkes <jancsika at yahoo.com>:
>>>
>>> It looks like a linker error that would affect most of the objects.  I
>>> don't think there's
>>> anything specific to mean.c causing the error-- it's probably in the
>>> build
>>> scripts.
>>>
>>> This is a likely culprit:
>>> "--enable-auto-import "/pd.dll"
>>>
>>> That's not where pd.dll lives.
>>>
>>> -Jonathan
>>>
>>>
>>> ________________________________
>>> From: Derek Kwan <derek.x.kwan at gmail.com>
>>> To: Jonathan Wilkes <jancsika at yahoo.com>; brbrofsvl at gmail.com;
>>> porres at gmail.com; l2ork-dev at disis.music.vt.edu
>>> Sent: Thursday, March 30, 2017 5:42 AM
>>> Subject: Re: [L2Ork-dev] updating cyclone
>>>
>>> Jonathan Wilkes <jancsika at yahoo.com> writes:
>>>
>>> Hello,
>>>
>>> I don't really have ready access to a Windows build environment and I
>>> haven't built much in Windows, so I think Alex and Matt (who I think do
>>> have machines with Windows installed) will be of more help than I can be
>>> in this particular situation. I have been looking over mean.c though and
>>> nothing really jumps out at me as being markedly different from the
>>> other source files. I did notice an "unsigned" qualifier in the struct
>>> with x_count that didn't have a type but I think that aliases to
>>> "unsigned int" anyways so that shouldn't be the issue (although I did a
>>> pull on Alex's repo to change it anyways for clarity's sake). Are there
>>> issues building any other object in Windows or is it just [mean]?
>>>
>>> Derek
>>>
>>>
>>> > Hi,
>>> > Not sure if the two of you are subscribed to the l2ork-dev list. I
>>> > sent an
>>> > inquiry there regarding cyclone build errors on Windows:
>>> >
>>> > http://disis.music.vt.edu/pipermail/l2ork-dev/2017-March/001235.html
>>> >
>>> > Thanks,
>>> > Jonathan
>>>
>>> >
>>>
>>> --
>>> Derek Kwan
>>> www.derekxkwan.com
>
>>>
>>>
>>>
>>
>>
>> _______________________________________________
>> L2Ork-dev mailing list
>> L2Ork-dev at disis.music.vt.edu
>> http://disis.music.vt.edu/listinfo/l2ork-dev
>
>
>


More information about the L2Ork-dev mailing list