History log of /seL4-l4v-10.1.1/isabelle/src/Pure/System/system_channel.ML
Revision Date Author Comments
# a0395a6e 03-Mar-2016 wenzelm <none@none>

discontinued polyml-5.3.0;

--HG--
rename : src/Pure/RAW/ml_debugger_polyml-5.6.ML => src/Pure/RAW/ml_debugger.ML
rename : src/Pure/RAW/ml_name_space_polyml-5.6.ML => src/Pure/RAW/ml_name_space.ML
rename : src/Pure/RAW/ml_profiling_polyml-5.6.ML => src/Pure/RAW/ml_profiling.ML


# 9e8988bc 20-Aug-2015 wenzelm <none@none>

precise BinIO, without newline conversion on Windows;


# fec5b85b 11-Jan-2015 wenzelm <none@none>

discontinued fifo channel, always use portable socket;


# 7201bc07 16-Oct-2013 wenzelm <none@none>

clarified workaround: problem is potential blocking of vacuous input (see also Poly/ML SVN 1874);


# a755b111 15-Oct-2013 wenzelm <none@none>

avoid empty input and its adverse effect on socket communication, e.g. output message getting "stuck" after vacuous update due to Session.update_options;


# c2be72f4 09-Jan-2013 wenzelm <none@none>

recovered buffered sockets from 11f622794ad6 -- requires Poly/ML 5.5.x;


# 2f509816 17-Oct-2011 wenzelm <none@none>

always use sockets on Windows/Cygwin;
discontinued special raw_dump facility;


# 2191e157 23-Sep-2011 wenzelm <none@none>

raw unbuffered socket IO, which bypasses the fragile BinIO layer in Poly/ML 5.4.x;


# ac86da98 22-Sep-2011 wenzelm <none@none>

abstract System_Channel in ML (cf. Scala version);
back to TextIO for fifo, which is more stable in Poly/ML 5.4.x;
explicit block buffering -- BinIO might be subject to old Poly/ML defaults;