removing older deprecated functions

[John W. Eaton]

When we started marking functions as deprecated and moving them to the
scripts/deprecated directory, I think we discussed keeping the
obsolete functions in the Octave distribution for 2 releases, then
dropping them.  Is there any objection to doing that now for the
following functions, which were marked as deprecated in version 3.0
and so should be removed in the 3.4 release?

  beta_cdf.m         f_rnd.m               is_symmetric.m   struct_contains.m 
  beta_inv.m         gamma_cdf.m           is_vector.m      struct_elements.m 
  beta_pdf.m         gamma_inv.m           isstr.m          t_cdf.m           
  beta_rnd.m         gamma_pdf.m           lognormal_cdf.m  t_inv.m           
  binomial_cdf.m     gamma_rnd.m           lognormal_inv.m  t_pdf.m           
  binomial_inv.m     geometric_cdf.m       lognormal_pdf.m  t_rnd.m           
  binomial_pdf.m     geometric_inv.m       lognormal_rnd.m  uniform_cdf.m     
  binomial_rnd.m     geometric_pdf.m       meshdom.m        uniform_inv.m     
  chisquare_cdf.m    geometric_rnd.m       normal_cdf.m     uniform_pdf.m     
  chisquare_inv.m    hypergeometric_cdf.m  normal_inv.m     uniform_rnd.m     
  chisquare_pdf.m    hypergeometric_inv.m  normal_pdf.m     weibcdf.m         
  chisquare_rnd.m    hypergeometric_pdf.m  normal_rnd.m     weibinv.m         
  clearplot.m        hypergeometric_rnd.m  pascal_cdf.m     weibpdf.m         
  clg.m              intersection.m        pascal_inv.m     weibrnd.m         
  com2str.m          is_bool.m             pascal_pdf.m     weibull_cdf.m     
  exponential_cdf.m  is_complex.m          pascal_rnd.m     weibull_inv.m     
  exponential_inv.m  is_list.m             poisson_cdf.m    weibull_pdf.m     
  exponential_pdf.m  is_matrix.m           poisson_inv.m    weibull_rnd.m     
  exponential_rnd.m  is_scalar.m           poisson_pdf.m    wiener_rnd.m      
  f_cdf.m            is_square.m           poisson_rnd.m  
  f_inv.m            is_stream.m           polyinteg.m    
  f_pdf.m            is_struct.m           setstr.m       

If anyone feels strongly that we should not simply delete these
functions, then how about creating an "obsolete" package and keeping
them there instead of in the Octave distribution itself?

jwe