Re: whitespace_in_literal_matrix?

[Dave Comer]

Hello John,

I am using 2.1.73. I downloaded the "Test version", and under Fedora 
Core 5, configured, compiled and installed. I found the following 
"Built-in Variable" (per the book) were not listed in the help (I've 
attached the list of built-in variables reported by the version below):

empty_list_elements_ok (p44)
propagate_empty_matricies (p45)

I also found the following variable had no effect (per my prior message):

whitespace_in_literal_matrix (p41)

These could be a documentation issue (more likely stupid user 
tricks), but thought I'd ask. I had assumed that the compiled version 
would be more targeted, but not necessarily less complete). None of 
this is a big deal, just was curious to make sure I didn't screw something up.

Dave

*** built-in variables:

DEFAULT_EXEC_PATH              octave_core_file_format
DEFAULT_LOADPATH               octave_core_file_limit
EDITOR                         octave_core_file_name
EXEC_PATH                      output_max_field_width
FFTW_WISDOM_PROGRAM            output_precision
IMAGEPATH                      page_output_immediately
INFO_FILE                      page_screen_output
INFO_PROGRAM                   print_answer_id_name
LOADPATH                       print_empty_dimensions
MAKEINFO_PROGRAM               print_rhs_assign_val
PAGER                          save_header_format_string
PS1                            save_precision
PS2                            saving_history
PS4                            sighup_dumps_octave_core
__kluge_procbuf_delay__        sigterm_dumps_octave_core
ans                            silent_functions
automatic_replot               split_long_rows
beep_on_error                  string_fill_char
completion_append_char         struct_levels_to_print
crash_dumps_octave_core        suppress_verbose_help_message
current_script_file_name       variables_can_hide_functions
debug_on_error                 warn_assign_as_truth_value
debug_on_interrupt             warn_associativity_change
debug_on_warning               warn_divide_by_zero
debug_symtab_lookups           warn_empty_list_elements
default_save_format            warn_fortran_indexing
echo_executing_commands        warn_function_name_clash
fixed_point_format             warn_future_time_stamp
gnuplot_binary                 warn_imag_to_real
gnuplot_command_axes           warn_matlab_incompatible
gnuplot_command_end            warn_missing_semicolon
gnuplot_command_plot           warn_neg_dim_as_zero
gnuplot_command_replot         warn_num_to_str
gnuplot_command_splot          warn_precedence_change
gnuplot_command_title          warn_reload_forces_clear
gnuplot_command_using          warn_resize_on_range_error
gnuplot_command_with           warn_separator_insert
gnuplot_has_frames             warn_single_quote_string
history_file                   warn_str_to_num
history_size                   warn_undefined_return_values
ignore_function_time_stamp     warn_variable_switch_label
max_recursion_depth            whos_line_format




At 11:42 PM 7/8/2006, you wrote:
> On  8-Jul-2006, address@hidden wrote:
>
> | I just received a copy of "GNU Octave Manual" (wanted to support the
> | project of course) and am working through the book. It appears that
> | the 'whitespace_in_literal_matrix' built-in variable is no longer
> | supported:
> |
> | octave:33> whitespace_in_literal_matrix="ignore"
> | whitespace_in_literal_matrix = ignore
> | octave:34> [1 2]
> | ans =
> |
> |   1  2
> |
> | octave:35>
> |
> | Can someone confirm this? I realize that it does not appear in the 
> help, just curious as to when it dissapeared.
>
> What version of Octave are you using?  If it is 2.9.6, then take a
> at the NEWS file that is distributed with the Octave sources.  If you
> installed a binary package, it should be with the other "extra
> documentation" files.  Debian puts it in the file
> /usr/share/doc/octave2.9/NEWS.gz.
>
> jwe