Changes to READMEs

git-svn-id: https://svn.code.sf.net/p/kaldi/code/trunk@1356 5e6a8d80-dfce-4ca6-a32a-6e07a63d50c8

egs/rm/README.txt

@@ -19,11 +19,7 @@ scripts at the moment.
       
   s2: This setup is experiments with pure hybrid system.
 
-  s3: This is "new-style" recipes.  We recommend to look here first, for
-      RM recipes.
-      However, the WSJ or Switchboard s3/ recipes are probably a better
-      place to look if you're trying to set up something on your own 
-      data, because they're more configurable.
+  s3: This is "new-style" recipes (now superseded by s5)
 
   s4: A recipe based on freely available subset of RM data, distributed by CMU
 

egs/swbd/README.txt

@@ -17,8 +17,7 @@ scripts for a sequence of experiments.
    data, just what is in the LDC corpus.  As a result there is a
    high-perplexity language model and small dictionary.
 
-  s5: This is the "new-new-style" recipe.  It's not yet finished.  It will
-    run up to a certain point.  All further work will be on top of this
-    style of recipe.
+  s5: This is the "new-new-style" recipe.  
+    All further work will be on top of this style of recipe.
 
 

egs/wsj/README.txt

@@ -21,12 +21,7 @@ scripts at the moment.
   s2: This setup is experiments with pure hybrid system as well 
       as with Tandem bottleneck feature system.
 
-  s3: This is the "new-style" recipe.   We recommend to look here first.
-      The recipe uses a subset of the algorithms in s1, but adds
-      cepstral mean subtraction (cms), and also uses more flexible, 
-      general scripts.
+  s3: This is the "new-style" recipe, now superseded by s5.
 
-  s5: This is the newer-style recipe.  It should run, although the RESULTS
+  s5: This is the newest-style recipe.  It should run, although the RESULTS
       file is not finalized.
-
-