[PD-dev] path for embedded examples

[Hans-Christoph Steiner]

On Apr 12, 2010, at 3:43 AM, Roman Haefeli wrote:

> On Mon, 2010-04-12 at 09:33 +0200, Frank Barknecht wrote:
>> On Sat, Apr 10, 2010 at 11:01:22PM +0200, Roman Haefeli wrote:
>>> On Sat, 2010-04-10 at 13:25 -0400, Hans-Christoph Steiner wrote:
>>>> Part of the idea of a libdir is to also include examples.   
>>>> Currently,
>>>> this is done using an 'examples' subfolder.  The problem with  
>>>> that is
>>>> the example patches won't automatically find the objects from the
>>>> libdir that they are embedded in. So I'm wondering what the best  
>>>> way
>>>> to handle this is.  Here are a couple ideas:
>>>>
>>>> - require [declare -path ..] for all example patches (works on any
>>>> branch of Pd)
>>>>
>>>> - instead of 'examples' folder, use mypatch-example.pd in the  
>>>> main dir
>>>> of the libdir (works without extra tricks)
>>>>
>>>> - use [import mylib] for all example patches (works even when the
>>>> example patch is saved elsewhere)
>>>
>>> - use [declare -stdpath extra/mylib] for all example patches  
>>> (works on
>>> any branch of Pd and works even when the example patch is saved
>>> elsewhere)
>>
>> I think, this doesn't work if "mylib" isn't installed or isn't  
>> installed in
>> "extra".
>
> True. But since we're talking about example patches _coming with_
> certain libraries, I think it does make sense to assume that the
> original library is installed, and installed in "extra".
> Or am I missing something here?


Yeah, [declare -stdpath extra/mylib] requires the library to be in the  
'extra' folder.  [declare -path ..] will work anywhere that library  
is, but won't work if the example is saved outside of the library  
folder.

Also, stdpath will add the library to the global path, right?  That  
seems like a bad idea.  That could change the objects that other  
patches get without any real warning.  It needs to be something that  
only applies to that particular patch.

.hc

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

Computer science is no more related to the computer than astronomy is  
related to the telescope.      -Edsger Dykstra