From 7601ef8eabc854709b5bdb312de9e7cee34a8c20 Mon Sep 17 00:00:00 2001 From: Bryan Jurish Date: Fri, 23 Jan 2009 21:50:38 +0000 Subject: + added branch pdstring+wchar/ from pdstring/ svn path=/trunk/externals/moocow/; revision=10606 --- pdstring+wchar/README.pod | 68 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 68 insertions(+) create mode 100644 pdstring+wchar/README.pod (limited to 'pdstring+wchar/README.pod') diff --git a/pdstring+wchar/README.pod b/pdstring+wchar/README.pod new file mode 100644 index 0000000..0f8d0bf --- /dev/null +++ b/pdstring+wchar/README.pod @@ -0,0 +1,68 @@ +=pod + +README for pd external package 'pdstring' + +Last updated for pdstring v0.06 + +=head1 DESCRIPTION + +The 'pdstring' package contains objects for converting to and +from (ASCII)-strings, represented as lists of floats. + +=head1 INSTALLATION + +Issue the following commands to the shell: + + cd PACKAGE-XX.YY (or wherever you extracted the distribution) + ./configure + make + make install + +=head1 BUILD OPTIONS + +The 'configure' script supports the following options, among others: + +=over 4 + +=item * --help + +Output a brief usage summary of the 'configure' script, +including a list of supported options and influential +environment variables. + +=item * --enable-debug , --disable-debug + +Whether to enable verbose debugging messages. +Default=no. + +=back + +=head1 ACKNOWLEDGEMENTS + +PD by Miller Puckette and others. + +Ideas, black magic, and other nuggets of information drawn +from code by Guenter Geiger, Larry Troxler, and iohannes m zmoelnig. + +=head1 KNOWN BUGS + +=head2 Memory Usage + +Encoding each byte of a string as its own float is shamefully +wasteful: it uses only 1 byte out of at least 3 which could be +losslessly used given ANSI/IEEE Std 754-1985 floats, not to +mention the remaining byte(s) (usually 1) of the float itself +or the (usually 4) bytes used for the a_type flag. +Unfortunately, Pd trims some floating point +precision in message boxes and in float atoms, so a truly +lossless float encoding for Pd would only be possible using +2 bytes per float (wasting 1/2 the space of the float itself), +and (to me), the memory saving such an encoding would provide +is just not worth the lack of transparency and additional +workload it would involve (but contact me if you want the code +anyways). + + +=head1 AUTHOR + +Bryan Jurish Emoocow@ling.uni-potsdam.deE -- cgit v1.2.1