On 11/10/2014 12:55 AM, Jonathan Wilkes via Pd-list wrote:
On 11/08/2014 01:05 AM, IOhannes zmölnig wrote:
Could you share some use-cases?
My own starting point was a regression test patch for loading all potentially loadable binaries which come with a particular pd installation package. (Well, I say regression test-- when I ran it for the first time it revealed a few objects which have probably always crashed Pd-extended.)
[...]
Since the directory that houses "extra" also houses "doc", I just went ahead and made a method for "libdir", which looked to point to parent directory of those two directories.
so the main purpose of the "libdir path" is indeed to find the "extra" path - i don't think the "doc" path is super-interesting, as these days libraries don't install to it (any more), so content under "doc/" is well defined.
but if "extra" is what you after, then i thik the valueS of "sys_staticpath" are more intersting, as these enumerate all default search paths ("all extra paths").
but maybe there's a better name for that method since it conflicts with the name for anl self-contained external library that adheres to Hans' standard format.
How about "pdinstalldir"? (prefixed 'pd' to avoid confusion with the "install" path of the patch.)
I don't understand what this means. What would the "install" path of a patch be?
well, just the path of the patch, but viewed from an application pov, where you ship an application written in Pd (as opposed to the Pd-application).
i'm aware that you have already added other ways to query *this* info.
I have the scope of the query in the name of the object: [pdinfo]. Currently I have [dir(---[pdinfo] pointing to the directory of the pd executable, but I guess that could be changed.
so this returns /usr/(local/)?bin on an ordinary linux installation?
Anway, what you are calling "pdinstalldir" would return the single directory I am currently calling "libdir". Is that correct?
yes. this was just a suggestion for an alternate name of "libdir" which does not require the user to know anything about filesystem hierarchy standards...
fgdsamr IOhannes