On 05/08/2016 02:28 PM, Winfried Ritsch wrote: > tried /home/winfried/tmp/test_load/searchme.pd and succeeded > tried /home/winfried/tmp/test_load/searchme.pd and succeeded > > Note the double suceeded lines.
this is known behaviour. the first line is printed when the "searchme" object is found to be an abstraction. the second line is printed when the abstraction is loaded. you will notice that for subsequent invocations of [searchme], you will only get a single "succeeded" line (because Pd already nows that "searchme" is an abstraction). you will also notice that for subsequent invocations of [searchme], you will *not* see the "failed" lines for externals (e.g. searchme.pd_linux). it might be a bit confusion, but i did not think it confusion enough to spend coding energy on it - after all it is a double line of success when everything goes right. > > I also notices double suceeded on loading externals: like the i guess "externals" here refers to "libraries", since the [initbang] object in question is really an abstraction. > [acre/acre/initbang] from deken installed acre lib (last lines): > > [...] > tried /home/winfried/pd-externals/acre/acre/initbang/initbang.l_ia64 and > failed > tried /home/winfried/pd-externals/acre/acre/initbang/initbang.pd_linux and > failed > tried /home/winfried/pd-externals/acre/acre/initbang.pd and succeeded up to this line, Pd did not know that "initbang" was an abstraction. now it does now it, and tries to instantiate it by loading the abstraction. > tried /home/winfried/tmp/test_load/acre/acre/initbang.pd and failed > tried /home/winfried/pd-externals/acre/acre/initbang.pd and succeeded since loading the abstraction involves finding it on disk, Pd iterates through the paths again. the two lines above demonstrate this. > > see double succeeded line > > If this an unkown behaviour I can have a look in the s_loader.c > what you are seeing is really just a side-effect of the new abstraction loader, which now uses the same mechanics as other loaders (eg. the external loader). you might think it is inefficient to search two times for a single file. however, in reality you get a performance boost as soon as you load an abstraction multiple times. e.g. when using [initbang] three times, the new loader will output: tried ~/tmp/test_load/acre/initbang/initbang.l_ia64 and failed tried ~/tmp/test_load/acre/initbang/initbang.pd_linux and failed tried ~/pd-externals/acre/initbang/initbang.l_ia64 and failed tried ~/pd-externals/acre/initbang/initbang.pd_linux and failed tried ~/tmp/test_load/acre/initbang.pd and failed tried ~/pd-externals/acre/initbang.pd and succeeded tried ~/tmp/test_load/acre/initbang.pd and failed tried ~/pd-externals/acre/initbang.pd and succeeded tried ~/tmp/test_load/acre/initbang.pd and failed tried ~/pd-externals/acre/initbang.pd and succeeded tried ~/tmp/test_load/acre/initbang.pd and failed tried ~/pd-externals/acre/initbang.pd and succeeded whereas the old loader would have done: tried ~/tmp/test_load/acre/initbang/initbang.l_ia64 and failed tried ~/tmp/test_load/acre/initbang/initbang.pd_linux and failed tried ~/pd-externals/acre/initbang/initbang.l_ia64 and failed tried ~/pd-externals/acre/initbang/initbang.pd_linux and failed tried ~/tmp/test_load/acre/initbang.pd and failed tried ~/pd-externals/acre/initbang.pd and succeeded tried ~/tmp/test_load/acre/initbang/initbang.l_ia64 and failed tried ~/tmp/test_load/acre/initbang/initbang.pd_linux and failed tried ~/pd-externals/acre/initbang/initbang.l_ia64 and failed tried ~/pd-externals/acre/initbang/initbang.pd_linux and failed tried ~/tmp/test_load/acre/initbang.pd and failed tried ~/pd-externals/acre/initbang.pd and succeeded tried ~/tmp/test_load/acre/initbang/initbang.l_ia64 and failed tried ~/tmp/test_load/acre/initbang/initbang.pd_linux and failed tried ~/pd-externals/acre/initbang/initbang.l_ia64 and failed tried ~/pd-externals/acre/initbang/initbang.pd_linux and failed tried ~/tmp/test_load/acre/initbang.pd and failed tried ~/pd-externals/acre/initbang.pd and succeeded gfmsard IOhannes
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Pd-dev mailing list [email protected] https://lists.puredata.info/listinfo/pd-dev
